Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator

Waste Management xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Waste Management journal homepage: www.elsevier.com/locate/wasman

Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator Gontran Richard a,b, Seddik Touhami a, Thami Zeghloul a, Lucien Dascalescu a,⇑ a b

PPRIME Institute, CNRS – Université de Poitiers – ENSMA, IUT, 4 avenue de Varsovie, Angoulême 16021, France CITF, Dorgeville, 16170 Saint-Cybardeaux, France

a r t i c l e

i n f o

Article history: Received 28 March 2016 Revised 11 June 2016 Accepted 28 June 2016 Available online xxxx Keywords: WEEE Electric wire wastes Design of experiments Electrostatic separation Electric field Particulate matter

a b s t r a c t Plate-type electrostatic separators are commonly employed for the selective sorting of conductive and non-conductive granular materials. The aim of this work is to identify the optimal operating conditions of such equipment, when employed for separating copper and plastics from either flexible or rigid electric wire wastes. The experiments are performed according to the response surface methodology, on samples composed of either ‘‘calibrated” particles, obtained by manually cutting of electric wires at a predefined length (4 mm), or actual machine-grinded scraps, characterized by a relatively-wide size distribution (1–4 mm). The results point out the effect of particle size and shape on the effectiveness of the electrostatic separation. Different optimal operating conditions are found for flexible and rigid wires. A separate processing of the two classes of wire wastes is recommended. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Society evolution leads to the generation of increased quantities of electric and electronic equipment wastes (WEEE) (Li et al., 2013). Decrease of fossil resources makes their recycling necessary. That is why a lot of studies have been conducted on the recovery of valuable materials from end of life computers, printers, fridges, TV-sets, cell-phones, and other such devices (Veit et al., 2005; Oguchi et al., 2011; Menad et al., 2013). Electric wires represent a significant part of these wastes, so it is important to find efficient and reliable solutions to recycle them (Bezerra de Araújo et al., 2008). The two major steps of the technologies that allow the recovery of metals and plastics from this kind of waste are: (1) finely-grinding of the scrap wires, to dissociate the constitutive materials; (2) separation of the constituents, based on the differences in their mass density and/or electric conductivity. Roll-type corona-assisted electrostatic separators are widely used for sorting out of conductive and non-conductive particles from a variety of mixed granular materials, such as minerals, industrial wastes or agricultural products (Ralston, 1961; Félici, 1966; Lawver and Dyrenforth, 1973; Haga, 1995; Dascalescu et al., 1998; Brands et al., 2000; Kohnlechner and Dascalescu, 2005; Tilmatine et al., 2009). In these installations, the electrical ⇑ Corresponding author. E-mail address: [email protected] (L. Dascalescu).

field, allowing electrostatic separation, is created between a rotating cylindrical electrode, connected to the ground, and one or more high-voltage electrodes. At least one of them generates a DC corona discharge (Dascalescu et al., 1995; Iuga et al., 2001, 2011). S-shaped plate-type electrostatic separator (Fig. 1) is another solution (Inculet et al., 1998; Dascalescu et al., 2001; Das et al., 2007). In this case, the electrical field is created between an elliptical-cross-section cylindrical electrode (1) that is connected to the high-voltage, and a grounded S-shaped metal plate (2). The only electric charging mechanism is the electrostatic induction (Vlad et al., 2000a,b; Park et al., 2015). Granules (more or less good conductors) are transported by the electromagnetic vibratory feeder (3) and deposited at the upper edge of the plate electrode (2). Then, they slide down along the surface of the plate and behave differently depending on their electrical conductivity. Conductive granules, by coming in the electric field area generated by highvoltage electrode (1), acquire by electrostatic induction an electric charge, the polarity of which is opposed to that of the high-voltage electrode. They are affected by three main forces: the electric attraction (Coulomb) force oriented to the elliptical electrode, the gravity force and the centrifugal force perpendicular to the surface of the S-shaped electrode (2). When the sum of the electric and centrifugal force surpasses the normal component of gravity force, the conductive granules detach from plate electrode and accumulate in the right-hand-side compartment of the collector (4). The non-conductive granules

http://dx.doi.org/10.1016/j.wasman.2016.06.036 0956-053X/Ó 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

2

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

2. Materials

Fig. 1. Laboratory plate-type electrostatic separator (CARPCO, Inc.); (1) ellipticalcross-section cylindrical electrode connected to the high-voltage supply (negative polarity); (2) grounded plate electrode; (3) electromagnetic vibratory feeder; (4) collector.

acquire no electric charge by electrostatic induction. They are not attracted by the high-voltage electrode (1) and the sum of the gravitational and centrifugal forces drives them in a different compartment. It is possible to add a corona electrode to better control non-conductive granules trajectories by charging them in an ionized electric field (Zeghloul et al., 2015). The reported applications of this separator have mainly been in the field of minerals separation (Iuga et al., 2004, 2011). The researchers have also performed numerical simulations of the electric field and of particles trajectories in these types of separators, for various electrode configurations (Vlad et al., 2000a,b, 2003; Abouelsaad et al., 2013). This is the simplest electrostatic separator that can be used for the selective sorting of conductive and non-conductive constituents of a granular mixture. It is characterized by less energy consumption and less maintenance that the roll-type electrostatic separators. The present work is aimed at validating a methodology that would optimize the outcome of an industrial plate-type electrostatic separation process employed for the selective sorting of granular metals and plastics, originating from either flexible or rigid electric wire wastes. Therefore, the two control factors that have been chosen are the ones that can commonly be adjusted in an industrial separator: the inclination angle of the high-voltage electrode and the high-voltage value applied on the elliptical electrode. These two factors act directly on the electric field and hence on the efficiency of the separation. The study is performed according to the response surface methodology (Frigon and Mathews, 1996; Goupy, 1999), on samples composed of either manually-cut ‘‘calibrated” cylindrical copper and PVC particles (length: 4 mm), or actual industrial grinded electric wire scraps, characterized by a relatively-wide size distribution (1–4 mm). The paper describes an experimental procedure that has paved the way from the laboratory study to industrial application. The electrostatic separator designed, engineered and manufactured in accordance with authors’ recommendations is capable of typically treating 100 kg of grinded WEEE per hour. The purity of the recycled copper is higher than 90%, for a recovery rate of 80% or more.

Flexible electric wires are composed of a multitude of copper filaments wrapped with a layer of PVC as electrical insulation. They are composed of roughly 55% of copper and 45% of PVC, in terms of mass. Conversely, rigid wires are made of a single strand of copper, surrounded with a layer of PVC. The percentage of copper is higher (about 65%) and the one of PVC is lower (35%) than in flexible wires. For a first set of experiments, primarily aimed at validating the feasibility of the electrostatic separation, ‘‘calibrated” cylindrical particles of 4 mm long are manually-cut from flexible and rigid electrical wire wastes having a cross-section of 0.5 mm2 (Figs. 2a and 3a). The other experiments are performed with machine-grinded electrical wire wastes of various sections. The plastic particles of this second mixture are characterized by a wide dispersion of size (1–4 mm) and various shapes (Figs. 2b and 3b). The presence of hybrid particles (i.e. unfree copper wire in plastic insulation – Fig. 3c) is expected to deteriorate the quality of electrostatic separation. The insulation hampers the contact between the copper wire and the plate electrode and hence its charging by electrostatic induction. These uncharged granules will not be recovered in the same compartment as the induction-charged copper ones. At the same time, they are too heavy to fall in the same compartment with the majority of the plastics. As a consequence, they are separately recovered with the middling product, which contains some copper granules deviated after an impact with the high-voltage electrode (Fig. 4).

3. Experimental procedure The experiments described hereafter are performed on a laboratory electrostatic separator (model EHTP(2 5,36)111
15, Carpco Inc., Jacksonville, Fl), the collector of which is composed of 16 compartments, each 25-mm wide (Fig. 5). These compartments are numbered, from right to left, from 1 to 16, and are grouped in 4 sections. The first one, B1, corresponds to compartments #1 to #7 and it principally collects the copper particles. Sections B2.1 and B2.2 correspond to compartments #8 and #9, in which are collected both copper and plastic (the former usually contains mainly copper, while the latter collects a majority of plastic particles). These two boxes are distinguished because, in some case, one (or both) of them could be pure enough to be considered as pure product. The product recovered in section B3 is composed of the previously-described ‘‘hybrid” granules, as well as a middling of copper and plastic particles which bounced off the other compartments. Compartments #14 to #16 are useless because no granules are recovered in them. To obtain optimum functioning point of the separator for the four sample types of granular electric wire wastes, the experiments are performed using the response surface method (Frigon and Mathews, 1996; Goupy, 1999; Hicks and Turner, 1999; Eriksson et al., 2000). This method has already been used to define optimum functioning point of other electrostatic separation processes (Dascalescu et al., 2004; Medles et al., 2007). The experimental design is focused on the variation of two factors (Fig. 5): – Inclination angle of the elliptical electrode (a, °). – High-voltage applied to the elliptical-cross-section cylindrical electrode (U, kV). Inclination angle of the plate electrode (b, °) is fixed at 55°, which has been defined as the optimum value by a series of

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

3

Fig. 2. (a) ‘‘Calibrated” granules, manually-cut from flexible electric wires (sample type #1); (b) granules originating from actual machine-grinded flexible electric wires (sample type #2).

Fig. 3. (a) ‘‘Calibrated” granules manually-cut from rigid electric wires (sample type #3); (b) granules originating from actual machine-grinded rigid electric wires (sample type #4); (c) hybrid particles (unfree copper wire in plastic insulation).

Fig. 4. Schematic representation of the plate-type electrostatic separation process; B1: copper product; B2: PVC product; B3: middling.

preliminary experiments. At lower values of b, the PVC granules no longer slide down along the S-shaped plate electrode. At b > 55°, the copper granules attain too high sliding speeds for the electric field forces to be able to appropriately control their trajectories after lift-off. For a given angle a, the increase of b is accompanied by a diminution of the electric field and hence of the action it exerts on the charged granules. The experimental domain of these two factors has been established after a series of preliminary experiments. The lower, central and upper values, designated respectively as 1, 0 and +1 levels, are given in Tables 1 and 2. No electric arc appears when the factors take values within these variation ranges, and a satisfactory

Fig. 5. Geometry of the laboratory plate-type electrostatic separator.

separation of copper and PVC granules is achieved (i.e., the purity of the copper product exceeds 85%). The voltage variation range is not the same for two types of wire wastes. For rigid wires, copper granules are heavier than those

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

4

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

formulas are used to calculate the recovery of mixed and PVC product and the PVC purity. Tables 5 and 6 summarize respectively the results obtained for the manually-cut and machine-grinded flexible electric wire wastes. Using the commercial software (MODDE 5.0, from Umetrics, Sweden), the recovery and purity of each product can be expressed as function of the centered normalized values xi of the n variables xi, i = 1, . . ., n, where:

Table 1 Variation ranges of the two control factors for the electrostatic separation of the granular flexible wire samples (#1) and (#2). Levels

a (°)

U (kV)

1 0 1

30 40 50

20 25 30

xi ¼ ðxi  xi0 Þ=Dxi ;

Table 2 Variation ranges of the two control factors for the electrostatic separation of the granular rigid wire samples (#3) and (#4). Levels

a (°)

U (kV)

1 0 1

30 40 50

30 35 40

ð3Þ

with:

xi0 ¼ ðximax þ ximin Þ=2; Dx ¼ ðximax  ximin Þ=2

ð4Þ

ximax and ximin being respectively the upper and lower limit of the experimental domain of the variable xi (Eriksson et al., 2000). In the present case, n = 2, and the two centered normalized variables are x1 ¼ U  , and x2 ¼ a . The MODDE 5.0 - calculated function that expresses the recovery of the Cu product as function of the variables U⁄ and a⁄ is:

from flexible ones, and this is why higher voltage levels are necessary. For each test the material flow rate is constant and equal to 100 g/min.

REC Cu ¼ 77:82 þ 22:93 U   0:45 a  8:36 U 2  1:62 a2 þ 0:89 U  a : ð5Þ

4. Results and discussion

Similar formulas are obtained for PuCu, RECPVC, PuPVC, as well as for the percentage represented by the mass of the middling product, collected in box B3, MASSB3. The values of the coefficients of the free term, as well as of U⁄, a⁄, U⁄2, a⁄2, and U⁄a⁄ for each of these functions are given in Tables 7 and 8, for respectively the manually-cut and the machine-grinded samples. The statistical indexes ‘‘goodness of fit” R2 and ‘‘goodness of prediction” Q2 are also calculated and listed in the above-mentioned tables. A good model is characterized by values of R2 and Q2 close to unity. The iso-response contour-plots for RECCu (%) and PUCu (%) of manually-cut ‘‘calibrated” flexible electric wire wastes are represented in Fig. 6. Similar plots obtained for the MassB3 (%), RECPVC (%) and PUPVC (%) are given in Figs. 7 and 8. The curves in the above-mentioned figures show that the inclination angle a has a quite weak effect on copper recovery, but strongly influences the recovery of the PVC product. Unsurprisingly, the voltage U has a strong positive effect on both the recovery and purity of the copper product. Similar curves are obtained and similar conclusions are derived for the machine-grinded wastes. MODDE 5.0 allows users to determine the optimum operating point of the process, which means the one where maximum percentages of copper and PVC are recovered, with maximum purities (Max RECCu, Max PUCu, Max RECPVC and Max PUPVC), while the mass of the middling should be minimal (Min MassB3). The coordinates of the optimal operating point in the case of manually-cut ‘‘calibrated” flexible electric wire wastes are:

4.1. Flexible electric wires With these variation ranges, an 11-run composite factorial experimental design is completed with samples #1 (Table 3) and #2 (Table 4): – 4 experiments at combinations of high and low levels of each factors (22), – 4 experiments at combinations of high and low levels of each factor and the average level of the other one (2 ⁄ 2), – 3 experiments at average level of both factors, to allow the software to evaluate the experimental error. In all these experiments, few PVC granules are collected in boxes B1 and B2.1. Therefore, the masses of the products collected in these boxes are considered together to represent the Cu product, the recovery and purity of which are calculated as follows:

mCu B1 þ mCu B2:1  100 mT Cu mCu B1 þ mCu B2:1 ¼  100 mT B1 þ mT B2:1

REC Cu ¼

ð1Þ

PU Cu

ð2Þ

where mCu Bi correspond to the mass of conductor collected in the compartment Bi, mT Cu is the total mass of conductor in all compartments and mT Bi is the total mass of the compartment Bi. Similar

Table 3 Results of the composite factorial experimental design carried out with 10-g samples (#1) of manually-cut ‘‘calibrated” flexible electric wire wastes. Mass (g) of granules collected in the boxes B1

B2.1

B2.2

B3

Voltage level

Angle level

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

1.05 2.89 1.19 2.83 1.26 2.84 1.64 2.14 1.93 2.21 2.16

0.93 2.78 1.01 2.63 1.08 2.70 1.49 1.91 1.78 2.01 1.88

0.12 0.1 0.19 0.18 0.17 0.18 0.15 0.23 0.15 0.19 0.28

2.39 2.47 1.98 2.2 2.06 2.17 3.14 2.5 2.49 2.8 2.95

2.06 2.07 1.61 2.02 1.7 1.9 2.76 2.24 2.18 2.52 2.58

0.33 0.37 0.37 0.18 0.35 0.27 0.38 0.25 0.3 0.26 0.33

5.33 2.65 4.85 2.37 4.53 2.18 3.36 2.81 3.3 3.07 2.76

2.29 0.53 3 0.55 2.6 0.7 1.34 1.24 1.09 0.68 0.69

3.04 2.12 1.85 1.82 1.93 1.48 2.02 1.57 2.21 2.39 2.07

1.16 1.99 2.07 2.57 2.1 2.77 1.83 2.5 2.3 1.88 2.09

0.48 0.33 0.35 0.31 0.49 0.42 0.29 0.35 0.32 0.29 0.61

0.68 1.66 1.72 2.26 1.61 2.35 1.54 2.15 1.98 1.59 1.48

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

5

G. Richard et al. / Waste Management xxx (2016) xxx–xxx Table 4 Results of the composite factorial experimental design carried out with 10-g samples (#2) of machine-grinded flexible electric wire wastes. Mass (g) of granules collected in the boxes B1

B2.1

B2.2

B3

Voltage level

Angle level

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

0.87 2.81 1.09 2.84 1.09 2.8 1.94 2.07 1.8 1.79 1.9

0.72 2.64 0.86 2.68 0.9 2.54 1.68 1.75 1.63 1.69 1.71

0.15 0.15 0.23 0.14 0.19 0.25 0.25 0.3 0.17 0.1 0.19

2.17 3.01 1.81 2.76 2.48 2.82 2.83 2.96 2.92 2.95 3.06

1.83 2.5 1.46 2.43 2.09 2.55 2.57 2.61 2.58 2.51 2.72

0.34 0.5 0.36 0.33 0.39 0.27 0.26 0.36 0.33 0.42 0.32

6.14 3.77 6.4 3.84 5.84 3.86 4.58 4.26 4.57 4.66 4.37

2.44 0.26 2.82 0.29 2.13 0.27 0.94 0.83 0.98 0.97 0.73

3.73 3.56 3.59 3.7 3.67 3.61 3.67 3.45 3.64 3.76 3.69

0.8 0.4 0.69 0.44 0.65 0.52 0.63 0.71 0.67 0.56 0.65

0.61 0.2 0.46 0.2 0.48 0.24 0.41 0.41 0.41 0.43 0.44

0.18 0.19 0.22 0.23 0.15 0.27 0.22 0.29 0.26 0.12 0.2

Table 5 Recovery and purity of the Cu and PVC products separated in the 11 runs of the composite factorial experimental design carried out with 10-g samples (#1) of manually-cut ‘‘calibrated” flexible electric wire wastes. Boxes B1 + B2.1 (Cu product) Voltage level 1 1 1 1 1 1 0 0 0 0 0

Angle level 1 1 1 1 0 0 1 1 0 0 0

RECCu (%) 51.91 84.94 43.89 84.39 47.36 80.41 72.28 72.30 73.74 82.36 77.43

PUCu (%) 86.92 90.49 82.65 92.45 83.73 91.74 88.91 89.44 89.59 90.42 87.28

Box B3 (middling) (%) of total mass 11.68 19.90 20.52 25.78 21.11 27.81 18.36 25.13 22.95 18.88 20.98

Box B2.2 (PVC product) RECPVC (%) 72.90 49.88 44.79 40.99 47.54 34.58 49.39 37.38 47.63 53.95 49.76

PUPVC (%) 57.04 80,00 38.14 76.79 42.60 67.89 60.12 55.87 66.97 77.85 75.00

Table 6 Recovery and purity of the Cu and PVC products separated in the 11 runs of the composite factorial experimental design carried out with 10-g samples (#2) of machine-grinded flexible electric wire wastes. Boxes B1 + B2.1 (Cu product) Voltage level

Angle level

RECCu (%)

PUCu (%)

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

45.54 91.79 41.43 91.25 49.39 90.89 75.89 77.86 75.18 75.00 79.11

83.88 88.32 80.00 91.25 83.75 90.57 89.10 88.68 89.19 88.61 89.31

Box B3 (middling) (%) of total mass 8.02 4.00 6.91 4.45 8.46 5.20 6.31 6.10 6.73 5.62 6.51

Box B2.2 (PVC product) RECCu (%)

PUCu (%)

84.77 80.91 81.59 84.09 83.41 82.05 83.41 83.41 82.73 85.45 83.86

60.75 94.43 56.09 96.35 58.84 93.52 80.13 80.99 79.65 80.69 84.44

– a = 30°, – U = 26.55 kV. For the machine-grinded flexible electric wire wastes, the optimum values of the two control variables of the separation process are: – a = 50°, – U = 30 kV.

Table 7 Values of the coefficients of the response functions RECCu, PUCu, MASSB3, RECPVC, PUCu, and of the statistical indexes R2 and Q2 for the composite factorial experimental design carried out with 10-g samples (#1) of manually-cut ‘‘calibrated” flexible electric wire wastes.

Free term U⁄

a

⁄

U⁄2

a⁄2 a⁄U⁄ Q2 R2

RECCu (%)

PUCu (%)

MassB3 (%)

RECPVC (%)

PUPVC (%)

74.53 17.76 2.09 9.59 0.81 1.87 0.945 0.991

88.82 3.56 0.63 1.47 0.97 1.56 0.795 0.956

21.76 3.36 3.75 2.91 5.30 0.74 0.967 0.990

48.87 6.63 9.50 7.98 11.34 4.80 0.988 0.997

71.63 14.48 5.39 17.03 8.72 3.92 0.908 0.980

Table 8 Values of the coefficients of the response functions RECCu, PUCu, MASSB3, RECPVC, PUCu, and of the statistical indexes R2 and Q2 for the composite factorial experimental design carried out with 10-g samples (#2) of machine-grinded flexible electric wire wastes.

Free term U⁄

a

⁄

U⁄2

a⁄2 a⁄U⁄ Q2 R2

RECCu (%)

PUCu (%)

MassB3 (%)

RECPVC (%)

PUPVC (%)

77.82 22.92 0.45 8.36 1.62 0.89 0.921 0.991

89.52 3.75 0.23 2.63 0.89 1.70 0.938 0.993

6.78 1.62 0.14 0.11 0.74 0.39 0.925 0.988

83.29 0.45 4.56E06 0.56 0.11 1.59 0.748 0.939

81.41 18.10 0.31 4.60 0.22 1.64 0.959 0.991

Optimum functioning point is different for manually-cut or machine-grinded flexible electric wire wastes. Tables 9 and 10 summarize the quantities of product that are recovered in each compartment for optimum functioning point of each type of waste. For both manually-cut ‘‘calibrated” and machine-grinded flexible electric wire wastes, the experimental results are in good agreement with MODDE 5.0 predictions. This validates the models used by the software. The separation efficiency of the copper product, calculated with the following formula:

gCu ¼

  mCu B1 þ mCu B2:1 mPVC B1 þ mPVC B2:1  100  mT Cu mT PVC

ð6Þ

is respectively 70.35% and 78.57%, for the manually-cut ‘‘calibrated” and machine-grinded electric wire wastes. The differences between these efficiencies are due to several factors. The main factor: the machine-grinded flexible electric wire wastes are characterized by non-homogeneous granule sizes. Many granules are smaller than the ‘‘calibrated” ones, and can be more easily deviated in the electric field. Another factor: the granules get charged during the

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

6

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

Fig. 6. MODDE 5.0 – computed iso-response contour plots of copper recovery and purity for manually-cut ‘‘calibrated” flexible electric wire waste (samples #1), for variable a and U.

Fig. 7. MODDE 5.0 – computed iso-response contour plots of middling mass for manually-cut ‘‘calibrated” flexible electric wire waste (samples #1), for variable a and U.

grinding process, by tribo-electric effect, while the manually-cut particles are electrically neutral when they arrive on the plate electrode. The residual charge of the machine-grinded wires can influence their trajectories. In the present case, the residual charge has

Fig. 8. MODDE 5.0 – computed iso-response contour plots of PVC recovery and purity for manually-cut ‘‘calibrated” flexible electric wire waste (samples #1), for variable a and U.

a beneficial effect on the separation: the PVC granules are negatively charged and they are repelled by the negative high voltage applied to the elliptical electrode. That is why, for machinegrinded wastes, the purity of recovered copper is higher than that obtained when processing manually-cut granules. The shape of the two classes of PVC granules can also explain their different behavior. Cylindrical calibrated granules get a higher speed when sliding along the grounded plate than the flatter machine-grinded ones. The higher initial speed of ‘‘calibrated” granules at the moment of their lift-off from the plate electrode may drive some of them into the boxes B1 or B2.1, reducing in this way the purity of the recovered copper product. Regarding copper purity, experimental results are better than the initial target of 90% at a recovery higher than 80%. The ‘‘calibrated” products just fulfill these requirements, but the performances recorded for the machine-grinded granules are significantly better. This observation points out the shape and size of the granules are of utmost importance for the efficiency of the electrostatic separation in this kind of equipment.

4.2. Rigid electric wires Two similar 11-run composite experimental designs have been performed, for respectively manually-cut ‘‘calibrated” and

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

7

G. Richard et al. / Waste Management xxx (2016) xxx–xxx Table 9 Predicted and experimental results at the optimum functioning point for the separation of manually-cut ‘‘calibrated” flexible electric wire wastes (#1). Boxes B1 + B2.1 (Cu product)

Predicted Experimental

Box B3 (middling) (%) of total mass

Voltage (kV)

Angle (°)

RECCu (%)

PUCu (%)

26.55 26.55

30 30

81.48 81.41

90.90 89.98

14.28 11.07

Box B2.2 (PVC product) RECPVC (%)

PUPVC (%)

65.34 76.51

87.38 87.14

Table 10 Predicted and experimental results at the optimum functioning point for the separation of machine-grinded flexible electric wire wastes (#2). Boxes B1 + B2.1 (Cu product)

Predicted Experimental

Box B3 (middling) (%) of total mass

Voltage (kV)

Angle (°)

RECCu (%)

PUCu (%)

30 30

50 50

91.21 86.98

91.22 93.00

4.55 5.20

Box B2.2 (PVC product) RECPVC (%)

PUPVC (%)

83.98 86.83

96.02 90.01

Table 11 Results of the composite factorial experimental design carried out with 10-g samples (#3) of manually-cut ‘‘calibrated” rigid electric wire wastes. Mass (g) of granules collected in the boxes B1

B2.1

B2.2

B3

Voltage level

Angle level

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

1.04 1.17 1.22 1.44 0.75 1.21 1.09 1.04 1.07 0.63 1.03

0.76 0.73 0.74 0.98 0.52 0.85 0.61 0.69 0.83 0.37 0.65

0.28 0.43 0.48 0.43 0.24 0.29 0.49 0.36 0.21 0.23 0.37

3.55 5.12 3.34 4.66 4.22 4.22 4.01 3.15 3.82 5.09 4.34

3.22 4.73 3.03 4.5 3.83 4.04 3.67 2.82 3.61 4.63 4.11

0.31 0.37 0.28 0.18 0.33 0.18 0.33 0.31 0.21 0.43 0.2

4.29 3.28 4.62 2.98 3.86 3.64 3.85 4.49 4.27 3.59 3.78

2.09 0.88 2.5 0.66 1.4 1.42 1.79 2.68 1.59 1.03 1.47

2.18 2.39 2.12 2.31 2.42 2.21 2.02 1.83 2.65 2.54 2.31

1.07 0.44 0.79 0.89 0.91 0.87 1.07 1.3 0.82 0.65 0.77

0.5 0.2 0.48 0.41 0.52 0.19 0.51 0.39 0.52 0.49 0.31

0.57 0.24 0.29 0.46 0.38 0.66 0.53 0.91 0.26 0.15 0.48

Table 12 Results of the composite factorial experimental design carried out with 10-g samples (#4) of machine-grinded rigid electric wire wastes. Mass (g) of granules collected in boxes B1

B2.1

B2.2

B3

Voltage level

Angle level

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

Total

Cu

PVC

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

0.94 0.78 1.16 1.23 0.83 1 0.85 0.83 1.03 0.92 0.8

0.7 0.63 0.87 0.91 0.68 0.73 0.68 0.63 0.85 0.73 0.53

0.24 0.14 0.25 0.29 0.14 0.25 0.17 0.2 0.19 0.18 0.28

4.61 4.5 3.31 4.29 3.91 4.75 4.32 5.73 4.97 4.51 4.73

3.74 4.11 2.95 3.84 3.31 4.31 3.95 5.05 4.29 3.92 4.43

0.85 0.38 0.31 0.44 0.6 0.45 0.35 0.66 0.67 0.6 0.3

2.84 2.64 3.8 2.37 3.41 2.15 3.08 2.27 2.1 2.62 2.53

1.34 1.08 1.69 0.95 1.54 0.83 1.35 0.41 0.63 1.01 0.67

1.49 1.53 2.02 1.42 1.3 1.3 1.73 1.84 1.48 1.61 1.87

1.64 2.1 1.76 2.16 1.83 2.02 1.68 1.22 1.81 1.88 1.81

0.42 0.39 0.57 0.48 0.64 0.26 0.16 0.15 0.38 0.5 0.5

1.18 1.72 1.19 1.65 1.17 1.75 1.52 1.06 1.42 1.37 1.3

machine-grinded rigid electric wire wastes. The experimental results obtained for the two classes of granular mixtures are presented in Tables 11 and 12. The respective calculated purity and recovery of the copper and PVC products of the electrostatic separation are listed in Tables 13 and 14. Based on these results, MODDE 5.0 software has computed the coefficients of the regression models that describe the responses of the separation process to the variation of the two control factors, a and U (Tables 15 and 16). The iso-response contour-plots for RECCu (%), PUCu (%), MassB3 (%), RECPVC (%) and PUPVC (%) of manually-cut ‘‘calibrated” rigid electric wire wastes are given in Figs. 9–11. The curves in Fig. 9 show

that the inclination angle a has a quite significant effect on both recovery and purity of the copper product. The electric field is more intense at the surface of the plate electrode when a = 40°. This factor also strongly influences the recovery of the PVC product (Fig. 11). As for the flexible electric wire wastes, the voltage U has a strong positive effect on both the recovery and purity of the copper product. The iso-response contour-plots obtained for the machine-grinded rigid electric wire wastes are slightly different, but the conclusions are similar to those derived for the manually-cut ‘‘calibrated” granules. The coordinates of the optimal operating point in the case of manually-cut ‘‘calibrated” rigid electric wire wastes are:

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

8

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

Table 13 Recovery and purity of the Cu and PVC products separated in the 11 runs of the composite factorial experimental design carried out with 10-g samples (#3) of manually-cut ‘‘calibrated” rigid electric wire wastes. Boxes B1 + B2.1 (Cu product) Voltage level

Angle level

RECCu (%)

PUCu (%)

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

60.58 83.49 55.85 70.66 69.38 89.23 65.05 53.34 67.79 76.69 72.78

86.71 86.8 82.68 89.84 87.53 92.06 83.92 83.77 88.8 88.41 88.64

Box B3 (middling) (%) of total mass 10.75 4.4 7.92 8.93 9.34 6.75 8.68 9.03 8.22 8.53 7.76

Box B2.2 (PVC product) RECPVC (%)

PUPVC (%)

65.27 69.68 66.88 68.34 71.81 75.17 59.94 60.67 69.58 68.82 68.75

50.82 72.87 45.89 63.52 62.69 78.71 52.47 40.76 62.06 65.75 61.11

Table 16 Values of the coefficients of the response functions RECCu, PUCu, MASSB3, RECPVC, PUCu, and of the statistical indexes R2 and Q2 for the composite factorial experimental design carried out with 10-g samples (#4) of machine-grinded rigid electric wire wastes.

Constant U⁄

a⁄ U⁄2

a⁄2 a⁄U⁄ Q2 R2

RECCu (%)

PUCu (%)

MassB3 (%)

RECPVC (%)

PUPVC (%)

77.3647 8.21667 1.895 3.97684 1.29816 5.0775 0.857 0.979

85.5095 1.805 0.371665 0.641319 0.198693 1.545 0.906 0.987

18.5695 1.095 0.508333 0.691316 1.12869 1.1675 0.917 0.979

41.2679 3.515 2.29833 4.05974 5.77026 4.29 0.836 0.966

70.65 9.43333 7.59333 20.91 0.409996 5.1825 0.823 0.977

Table 14 Recovery and purity of the Cu and PVC products separated in the 11 runs of the composite factorial experimental design carried out with 10-g samples (#4) of machine-grinded rigid electric wire wastes. Boxes B1 + B2.1 (Cu product) Voltage level

Angle level

RECCu (%)

PUCu (%)

1 1 1 1 1 1 0 0 0 0 0

1 1 1 1 0 0 1 1 0 0 0

70.61 76.33 62.83 88.86 64.67 82.22 75.41 82.03 78.58 75.49 77.91

83.00 89.77 85.46 86.05 84.18 87.65 85.56 84.59 85.67 85.64 85.69

Box B3 (middling) (%) of total mass 16.35 20.96 17.55 17.49 18.34 20.36 17.92 17.14 18.26 18.93 18.34

Box B2.2 (PVC product) RECPVC (%)

PUPVC (%)

39.63 40.58 53.58 37.37 40.5 34.67 45.89 48.94 39.36 42.82 40.87

30.46 57.95 53.16 59.92 38.12 60.47 60.17 81.06 70.48 68.45 73.91

Table 15 Values of the coefficients of the response functions RECCu, PUCu, MASSB3, RECPVC, PUCu, and of the statistical indexes R2 and Q2 for the composite factorial experimental design carried out with 10-g samples (#3) of manually-cut ‘‘calibrated” rigid electric wire wastes.

Free term U⁄

a⁄ U⁄2

a⁄2 a⁄U⁄ Q2 R2

RECCu (%)

PUCu (%)

MassB3 (%)

RECPVC (%)

72.0905

88.3042

8.32368

68.4611

9.595 4.87833 7.70869 12.4013 2.02499 0.902 0.963

1.96334 0.190001 1.95948 3.99053 1.7675 0.84 0.978

1.32167 0.341667 0.50921 0.300789 1.84 0.858 0.972

1.53834 0.166667 5.91237 7.27263 0.737498 0.82 0.972

PUPVC (%) 62.1453 9.28334 4.33167 9.79684 14.2882 1.105 0.834 0.974

– a = 38°, – U = 40 kV. For the machine-grinded rigid electric wire wastes, the optimum values of the two control variables of the separation process are: – a = 50°, – U = 36 kV. Experiments at these optimum functioning points gave the results which are presented in Tables 17 and 18.

Fig. 9. MODDE 5.0 – computed iso-response contour plots of copper recovery and purity for manually-cut ‘‘calibrated” rigid electric wire waste (samples #3), for variable a and U.

The copper efficiency separation of both types of granules is almost the same (about 67%) and the predicted results are close to the experimental ones. The copper recovery and purity are higher than 85%. However, the purity is slightly inferior to the target of 90%. Copper separation of rigid electric wires needs high values of the applied voltage U. The test bench used to make these experiments is limited to 40 kV but it can be expected that higher values of U give better separation results. In case of the machine-grinded rigid electric wire wastes, it is possible to increase PVC recovery and purity by adding B2.2

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

9

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

Fig. 10. MODDE 5.0 – computed iso-response contour plots of middling mass for manually-cut ‘‘calibrated” rigid electric wire waste (samples #3), for variable a and U.

compartment and B3 compartment (Table 19). In this case PVC recovery is of 80.41% and its purity is of 78.80%, which is better than the results obtained with ‘‘calibrated” granules, and there is no middling product. The cylindrical shape of ‘‘calibrated” PVC particles enables them to get a higher speed when sliding along the plate electrode and to be collected in B2.2. Grinded PVC particles, which are flatter, attain a lower speed and fall in B3. This fact explains the difference between the behavior of ‘‘calibrated” and machine-grinded PVC granules. The laboratory studies paved the way to the design and engineering of an industrial plate-type separator, Select CP 100, manufactured by CITF, France (Fig. 12). This system could treat 100 kg/h of electric wires wastes, grinded to a size of less than 3 mm. An electromagnetic feeder introduces the granular mixture in the separator at a constant rate. The middling product, which represents about 20% of the feed (rigid electric wire wastes in the example given in Fig. 13) and are mostly composed of hybrid granules, as the ones in Fig. 3c, is re-directed, by means of a screw, back in the grinder. After a second pass through the grinder, the probability of physically dissociating copper and insulation of hybrid granules highly increases. Grinding also reduces the size of some of the copper or PVC granules that have been collected with the middling product because they were too heavy for either being attracted to the high voltage electrode by the Coulomb force, or pinned to the plate electrode by the electric image force. With

Fig. 11. MODDE 5.0 – computed iso-response contour plots of PVC recovery and purity for manually-cut ‘‘calibrated” rigid electric wire waste (samples #3), for variable a and U.

the scheme in Fig. 13, the granules may pass through the electrostatic separator several times, until they are recovered in one of the two products, copper or PVC. Several open-loop tests (i.e., without re-treatment of the middling product) were conducted on the prototype of the industrial

Table 17 Predicted and experimental results at the optimum functioning point for the separation of manually-cut ‘‘calibrated” rigid electric wire wastes (#3). Boxes B1 + B2.1 (Cu product)

Predicted Experimental

Box B3 (middling) (%) of total mass

Voltage level

Angle level

RECCu (%)

PUCu (%)

40 40

38.08 38.08

90.26 86.16

91.78 89.31

6.09 6.58

Box B2.2 (PVC product) RECPVC (%)

PUPVC (%)

75.75 70.18

81.74 81.43

Table 18 Predicted and experimental results at the optimum functioning point for the separation of machine-grinded rigid electric wire wastes (#4). Boxes B1 + B2.1 (Cu product)

Predicted Experimental

Box B3 (middling) (%) of total mass

Voltage level

Angle level

RECCu (%)

PUCu (%)

36 36

50 50

83.06 87.44

85.02 87.51

16.95 17.02

Box B2.2 (PVC product) RECPVC (%)

PUPVC (%)

47.61 42.03

78.67 73.35

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

10

G. Richard et al. / Waste Management xxx (2016) xxx–xxx

Table 19 PVC recovery and purity for the experiment performed with machine-grinded rigid electric wire wastes (#4).

Box B3 Boxes B2.2 + B3

2.0 m

RECPVC (%)

PUPVC (%)

38.38 80.41

84.78 78.80

2.4 m

copper product is low (85%), as many PVC granules behave as conductors. At lower values of the relative humidity (64.5%), the purity of the copper product is excellent (97%). In both cases, the recovery of copper was beyond the target (85%). The main problem for an industrial application of the plate-type electrostatic separator is the dust which is created when wires are grinded. This dust pins on the electrodes and decreases the separation efficiency, as it hampers granule sliding down the plate electrode. A de-dusting operation should be included in the flowchart, prior to electrostatic separation. The studies performed on product de-dusting prior to electrostatic separation have not yet conducted to fully satisfactory results. The existing equipment operates well for de-dusting granular rigid cable waste, but is cannot be employed with the granular flexible cable waste, as it also removes a quite large quantity of fine copper particles. 5. Conclusions

2.4 m

Fig. 12. Photograph of the industrial plate-type electrostatic separator Select CP 100 manufactured by CITF, France.

Plate-type electrostatic separation is a complex multi-factorial process. Optimization of the outcome of an industrial process is far from being straightforward. The paper points out that the design of experiments methodology might facilitate such an action, and that the efficiency of plate-type electrostatic separators can be significantly improved by appropriate conditioning of the granular wastes that are fed to them. (1) The PVC granules of machine-grinded electric wire wastes carry a residual electric that affect their behavior in the electrostatic separator. This residual charge may be an advantage if its polarity is known, as the trajectories of PVC granules can be controlled by connecting the high-voltage electrode at a power supply of same polarity. Otherwise, it is important to use an AC corona electrode to neutralize this residual charge. This solution has been adopted for the Select CP 100 industrial separator designed by the authors. (2) The effectiveness of the electrostatic separation depends on the granule size and shape. Smaller and flatter granules are easier to control using the electrical and the mechanical forces in a plate-type separator. (3) High ambient humidity can deteriorate the outcome of the separation. Humid plastic granules behave as conductors and are collected in the copper product, the purity of which may be seriously reduced. (4) Experimental design methodology is a useful tool for modeling and optimization of the separation process for the various classes of electric wire wastes. (5) Different optimal operating conditions are found for the granular mixtures originating from flexible and rigid electric wire wastes. A separate processing of the two classes of wire wastes is recommended.

Fig. 13. Product flow in an industrial electrostatic separation facility.

electrostatic separator, using 100 kg of machine-grinded electric wire wastes. The Select CP 100 electrostatic separator can treat both flexible and rigid cables with a maximum section of 16 mm2. In spite of the fact that the optimum values of the control factors for flexible and rigid electric wire wastes are close to each other, the operation point is set in accordance to the type of product to be treated. Intermediate values can be adopted whenever the waste to be treated includes both types of electric wire wastes, but the performance indexes are altered. For rigid wires, the results of one-pass separation were fullysatisfactory: 95% of the copper were recovered in a conductive product having a purity of 90%. For flexible wires, the results depended a lot on humidity. If the humidity is high (83%) the purity of the

Acknowledgment The authors acknowledge with thanks their collaborators Cédric LUBAT, Maxime LUCAS and Frederic RAVAUD (CITF Company, SaintCybardeaux, France) for their support and experimental help. Thanks to them experiments could be made on an industrial facility. References Abouelsaad, M.M., Abouelatta, M.A., Salama, A.R., 2013. Genetic algorithmoptimised charge simulation method for electric field modelling of plate-type electrostatic separators. IET Sci. Meas. Technol. 7, 16–22. Bezerra de Araújo, M.C.P., Chaves, A.P., Espinosa, D.C.R., Tenório, J.A.S., 2008. Electronics scraps – recovering of valuable materials from parallel wire cables. Waste Manage. 28, 2177–2182.

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036

G. Richard et al. / Waste Management xxx (2016) xxx–xxx Brands, L., Beier, P.-M., Stahl, I., 2000. Electrostatic separation. In: Ullmann’s Encyclopedia of Industrial Chemistry. Wiley, New York. Das, S., Medles, K., Younes, M., Dragan, C., Dascalescu, L., 2007. Separation of fine granular mixtures in S-plate-type electrostatic separators. IEEE Trans. Ind. Appl. 43, 1137–1143. Dascalescu, L., Mizuno, A., Tobazéon, R., et al., 1995. Charges and forces on conductive particles in roll-type corona-electrostatic separators. IEEE Trans. Ind. Appl. 31, 947–956. Dascalescu, L., Morar, R., Iuga, A., Samuila, A., Neamtu, V., 1998. Electrostatic separation of insulating and conductive particles from granular mixes. Part. Sci. Technol. 16, 25–42. Dascalescu, L., Tilmatine, A., Aman, F., Mihailescu, M., 2004. Optimization of electrostatic separation processes using response surface modelling. IEEE Trans. Ind. Appl. 40, 53–59. Dascalescu, L., Vlad, S., Iuga, A., Levin, P.L., 2001. Electrostatics of conductive particles in contact with a plate electrode affected by a non-uniform electric field. J. Phys. D, Appl. Phys. 34, 60–67. Eriksson, L., Johansson, E., Kettaneh-Wold, N., Wikström, C., Wold, S., 2000. Design of Experiments. Principles and Applications. Learnways AB, Umea, Sweden. Félici, N.J., 1966. Forces et charges de petits objets en contact avec une électrode affectée d’un champ électrique. Rev. Gén. Elect. 75, 1145–1160. Frigon, N.L., Mathews, D., 1996. Practical Guide to Experimental Design. Wiley, New York, USA. Goupy, J., 1999. Experimental Design for Response Surfaces (in French, Plans D’expériences Pour Surfaces De Réponse). Dunod, Paris, pp. 253–255. Haga, K., 1995. Applications of the Electrostatic Separation Technique. Handbook of Electrostatic Processes. Marcel Dekker, pp. 365–368. Hicks, C.R., Turner Jr., K.V., 1999. Fundamental Concepts in the Design of Experiments. Oxford University Press, Oxford. Inculet, I.I., Castle, G.S.P., Brown, J.D., 1998. Electrostatic separation of plastics for recycling. Part. Sci. Technol. 16, 91–100. Iuga, A., Morar, R., Samuila, A., Dascalescu, L., 2001. Electrostatic separation of metals and plastics from granular industrial wastes. IEE Proc.-Sci. Meas. Technol. 148, 47–54. Iuga, A., Samuila, A., Neamtu, V., Morar, R., Belaca, R., Das, S., Dascalescu, L., 2011. Removal of metallic particles from acrylonitrile butadiene styrene wastes using electrostatic separation methods. IEEE Trans. Ind. Appl. 47, 322–329.

11

Iuga, A., Vlad, S., Mihailescu, M., Dascalescu, L., 2004. A laboratory plate/screen-type electrostatic separator for granular mixtures: design, engineering, and application. Part. Sci. Technol. 22, 275–283. Kohnlechner, R., Dascalescu, L., 2005. New applications for ‘‘standard” electrostatic separators. Conf. Rec. IEEE IAS Annu. Meet., 2569–2572 Lawver, J.E., Dyrenforth, W.P., 1973. Electrostatic Separation. A.D. Moore, pp. 221– 249. Li, J., Brenda, N., Lopez, N., Liu, L., Zhao, N., Yu, K., Zheng, L., 2013. Regional or global WEEE recycling. Where to go? Waste Manage. 33, 923–934. Medles, K., Tilmatine, A., Miloua, F., Bendaoud, A., Younes, M., Rahli, M., Dascalescu, L., 2007. Set point identification and robustness testing of electrostatic separation processes. IEEE Trans. Ind. Appl. 43, 618–626. Menad, N., Guignot, S., van Houwelingen, J.A., 2013. New characterization of electrical and electronic equipment wastes (WEEE). Waste Manage. 33, 706–713. Oguchi, M., Murakami, S., Sakanakura, H., Kida, A., Kameya, T., 2011. A preliminary categorization of end-of-life electrical and electronic equipment as secondary metal resources. Waste Manage. 31, 2150–2160. Park, C.-H., Subasinghe, N., Jeon, H.-S., 2015. Separation of covering plastics from particulate copper in cable wastes by induction electrostatic separation. Mater. Trans. 56, 1140–1143. Ralston, O.C., 1961. Electrostatic Separation of Mixed Granular Solids. Elsevier. Tilmatine, A., Medles, K., Bendimerad, S.E., Boukholda, F., Dascalescu, L., 2009. Electrostatic separators of particles: application to plastic/metal, metal/metal and plastic/plastic mixtures. Waste Manage. 29, 228–232. Veit, H.M., Diehl, T.R., Salami, A.P., Rodrigues, J.S., Bernardes, A.M., Tenório, J.A.S., 2005. Utilization of magnetic and electrostatic separation in the recycling of printed circuit boards scrap. Waste Manage. 25, 67–74. Vlad, S., Iuga, A., Dascalescu, L., 2000a. Modelling of conducting particle behaviour in plate-type electrostatic separators. J. Phys. D: Appl. Phys. 33, 127–133. Vlad, S., Iuga, A., Dascalescu, L., 2003. Numerical computation of conducting particle trajectories in plate-type electrostatic separators. IEEE Trans. Ind. Appl. 39, 66–71. Vlad, S., Mihailescu, M., Rafiroiu, D., Iuga, A., Dascalescu, L., 2000b. Numerical analysis of the electric field in plate-type electrostatic separators. J. Electrostat. 48, 217–229. Zeghloul, T., Dascalescu, L., Touhami, S., Miloudi, M., Dahou, O., Richard, G., 2015. Corona-assisted plate-type electrostatic separation process for granular plastic wastes. In: Conf. Rec. IEEE/IAS Annual Meeting, Dallas, October 18–22 2015.

Please cite this article in press as: Richard, G., et al. Optimization of metals and plastics recovery from electric cable wastes using a plate-type electrostatic separator. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.06.036