ARTICLE IN PRESS. Pulsed Electric Field Assisted Pressing of Sugar Beet Slices: towards a Novel Process of Cold Juice Extraction

Transcription

1 Biosystems Engineering (2006) 93 (1), doi: /j.biosystemseng PH Postharvest Technology ARTICLE IN PRESS Pulsed Electric Field Assisted Pressing of Sugar Beet Slices: towards a Novel Process of Cold Juice Extraction A.B. Jemai; E. Vorobiev Equipe des Technologies Agro-Industrielles, Département de Génie Chimique, Université de Technologie de Compiègne, Centre de Recherche de Royallieu, B.P , Compiègne, France; of corresponding author: eugene.vorobiev@utc.fr; Abdelbasset.Bessadok@issatgb.rnu.tn (Received 17 April 2004; accepted in revised form 14 September 2005; published online 28 November 2005) The use of pulsed electric fields (PEF) as an intermediate treatment for the cold juice extraction from sugar beet cossettes (i.e. long grated particles) has been investigated using a pilot scale multi-plate and frame pressing equipment (pressure of 5 15 bar; particles filling of kg) and a pulse generator (1000 V 1000 A). It has been possible to validate laboratory-scale results (40 g of particles) by studying the feasibility and advantages of a PEF-assisted cold pressing of sugar beet cossettes on a much larger scale (45 15 kg). A best scheme scenario for an adequate PEF-assisted cold pressing of sugar beet cossettes consists of two initial pressing steps with an intermediate PEF treatment, followed by one or more washing steps and a final pulp pressing. A yield of about 80% in juice per initial mass of cossettes has been achieved before washing. By performing some washing and final pressing operations, losses of sugar in pulp could be significantly reduced to about 3% of the initial sugar content. The purity of pressed juices was systematically higher following PEF treatment compared to that of juices prior to PEF treatment (96 98% and 90 93%, respectively). Spectrophotometric colour measurements, reflecting impurities and fine suspended particles contents in juices, showed that purified PEF juices had colour values three to four times lower than those of purified factory juices, a difference confirmed by the colour of sugar crystals obtained from both types of juices. In addition, significant amounts of potassium, sodium and a-amino nitrogen were found to remain in PEF-treated particles, which explains why better purity juices are obtained following PEF treatment. r 2005 Silsoe Research Institute. All rights reserved Published by Elsevier Ltd 1. Introduction The use of pulsed electric fields (PEF) to assist in the non-thermal sugar beet pressing process presents a real opportunity for the sugar industry. Since the late 1990s, the Industrial Agro-Food Technologies research unit (TAI) of the Compie` gne University of Technology (France) has been conducting laboratory-scale research work for the industrial implementation of this novel sugar extraction process (Bouzrara & Vorobiev, 2000, 2001; Bouzrara, 2001; Vorobiev et al., 2000). Other work dealing with PEF utilisation for sugar extraction has also been reported (Eshtiaghi & Knorr, 2002). To be fully accepted, this technology should be applicable to a larger scale with better process performances. Several laboratory-scale studies (i.e. 40 g of grated particles termed cossettes ) demonstrated the efficiency of applying a short duration PEF treatment during a two step cold pressing of sugar beet cossettes (Bouzrara & Vorobiev, 2000, 2001; Bouzrara, 2001). In general, the PEF was applied after a first period of pressing when extraction is significantly reduced due to cake compactness and blockage of drainage channels. Applying PEF at this particular instant increases the permeability of most of the remaining cell membranes leading to enhanced extraction of intra-cellular juice and better juice quality. Bouzrara and Vorobiev (2000, 2001) thus reported that up to around 82% of overall yield could be achieved by intermediately applying a PEF during a two stage pressing process at bar. At a pressure of /$ r 2005 Silsoe Research Institute. All rights reserved Published by Elsevier Ltd

2 58 A.B. JEMAI; E. VOROBIEV 5 bar, the yield was about 78%. In addition, the second juices (i.e. after PEF application) are systematically more concentrated in sugar and have lighter colour. Despite these encouraging results, questions as to whether or not the technology is transposable on an industrial scale remained unanswered. For viability, the technique has to be applicable to a much larger scale (e.g. from the semi-industrial to the industrial scale). In other words, it should fit into an overall extraction process, which at least quantitatively equals the conventional process and provides juices and pulps of better characteristics. The objectives of the present work were, first, to validate laboratory scale results on a larger scale (45 15 kg of cossettes) and then to demonstrate the benefits of using a PEF on the quantitative and qualitative levels on such a scale. 2. Materials and methods 2.1. Sugar beet Raw sugar beet roots were provided by a nearby sugar processing company (Be ghin Say, Chevrie` re, France). For most of this study, beets were received clean from the reception and control centre of the factory on a daily basis. During the final part of the study however, beets had to be stored unclean in a controlled temperature environment (3 5 1C) in a nearby vegetable storage company Equipment Plate and frame filter-press The filter-press (Choquenet, France) used throughout this study is illustrated by Fig. 1. It is a multi-plate and frame made of a number of membrane plates assembled one against the other, which are used to transmit pressure (5 15 bar) to the beet particles (45 15 kg). The dimensions of the plates are 470 mm by 470 mm with a cross-sectional area of cake of about 1370 cm 2. In most runs, one or two chambers were manually filled with grated cossettes (see Fig. 2 for shapes and cross-sections of two types of cossettes). In a few runs, up to six chambers were used at the same time. Each chamber consists of a plate covered by a filter cloth and a flexible electrode (metallic grid) on one side and a rigid electrode on the other. The pressure (compressed air) is applied to the membrane of the plate, which in turn exerts and distributes the pressure over the cossettes placed between the plate and the rigid electrode (Fig. 2). Juice is drained through channels leading to the outlet, where juice accumulation is monitored by a weighing balance connected to a data acquisition system Electrical and data acquisition systems The pilot scale generator (1000 V 1000 A) used throughout the present work was built by Hazemeyer Co. (Gauchy, France), who scaled up the equipment based on laboratory results. The ratings of the square pulse generator and the characteristics of the pulses are given in Table 1(a) and (b) Fig. 1. Simplified schematic of filter-press equipment (Bouzrara, 2001): (1) membrane plates; (2) and (3), false plates; (4) moving head; (5) fixed head; (6) guiding rail; (7) tightening piston; (8) filter body; (9) liquid recipient

3 PULSED ELECTRIC FIELD ASSISTED PRESSING OF SUGAR BEET SLICES 59 Porous flexible electrode Cossettes Grated type Pulse generator Porous rigid electrode Flexible membrane Industrial type Pressure Juice filtrate Filter cloth Juice evacuation channels Fig. 2. Schematic of a one chamber configuration of plate and frame filter press Table 1 Electrical equipment parameters and characteristics (a) Rating of the pulse generator Regulation control Maximum current, A An acquisition programme written in the HPVEE environment (HP-VEE, V312, Hewlett-Packard Co., USA) was used to control and command the generator and the weighing system Experimental procedure Maximum voltage, V Nominal power, kw Voltage (b) Pulse parameters Pulse type Period, ms Impulse duration, ms Frequency, Hz Uni-polar square Throughout the present study, more than 100 experiments have been performed; these were organised so as to explore several objectives such as the effects of pressure, the optimisation of washing operations, electrical power consumption, etc Cossette types and preparation In most runs, grated cossettes (20 mm in length and 4 mm in width) using a heavy-duty cutting equipment (Robo-coupe, France) were obtained by first cutting beetroots into 20 mm cubes, then grinding down the cubes using a grating disc of 4 mm mesh. To study the effect of cossette type, industrial cossettes (i.e. fine long particles of V-shaped cross-section) were obtained using a slicing equipment (provided by British Sugar), while the effect of varying the size was studied on grating type cossettes (e.g. 10 mm by 15 mm). Initial dry weight content d W in % (wet basis), sugar concentration S C in 1S (kg of sugar per 100 kg of solid), and soluble solids content S S in 1Brix (kg /100 kg juice) were systematically performed for each run. The raw beets had a dry weight of about (222707)%, a sugar concentration of ( ) 1S, and soluble solids content of ( ) 1Brix. To estimate the dry weight, 20 g of freshly cut cossettes were placed in an oven at 105 1C for 24 h; sugar concentration in 1S, soluble solids content in 1Brix, spectrophotometric colour in ICUMSA units (i.e. International Commission for Uniform Methods of Sugar Analysis), and juice handling and analysis were conducted by a Be ghin-say technician on site throughout the entire study, using sugar industry standard equipment and methods courtesy of Be ghin-say (France) Overall process steps The flow diagram of the process steps is illustrated by Fig. 3. In general, a PEF-assisted pressing process includes a first pressing step (1st press), a short time PEF treatment, a second pressing step (2nd press), and subsequent washing and pressing of pulps (1st wash, 2nd wash, etc.).

4 60 A.B. JEMAI; E. VOROBIEV Beets Slicing Grinding 1st press 1st juice Cossettes Filling by hand PEF and 2nd press 3rd Press 2nd juice 3rd juice Water 3 10% or 5 5% of initial mass of fresh cossettes Pressed cake Washing of cake and final pressing (1 or more steps) Final pulp 4th 3rd 2nd 1st wash wash wash wash Fig. 3. Experimental procedure: PEF, pulsed electric field It should be noted, however, that due to its actual construction, the plate and frame press used throughout this study did not allow complete pressing of the whole cake (outer cake ring not well pressed). Consequently, following the second pressing, the cake was systematically redistributed and pressed at the same pressure a third time (3rd press). So, pressed juices after PEF treatment include that of the 2nd and 3rd press. Following the 3rd press operation, the final cake was washed and pressed once or more frequently with fresh water (10% of initial weight of fresh cossettes). The latter was gradually poured over the pressed cake while mixing the particles to insure best possible water contact. The washed cake was then redistributed in one chamber of the filter press and re-pressed to obtain the wash liquor Optimisation methods Ways to optimise the process were tested by (i) varying the electrical energy input of the PEF treatment, (ii) exploring the washing steps, and (iii) increasing the applied pressure. From a power consumption stand point, the optimisation procedure consisted of studying the effect of reducing the frequency of the PEF (i.e. the number of impulses, which is proportional to the treatment time) on the juice yield. To do so, impulse numbers of 100, 300, 500, 750, and 1000 have been used, while keeping the applied potential difference constant (i.e V). In a cold pressing process, such as the one concerned with the present study, one or more washing steps were inevitable. In fact, significant juice and sugar amounts are always trapped in the pulp following the pressing operations. These substances may only be extracted by subjecting the pulp to a concentration difference using fresh water for example. The washed particles may then be re-pressed to obtain wash juices containing sugar. Two washing scenarios were studied for this purpose: (i) the first consisted of performing three washing steps using an amount of pure water equivalent to 10% of the initial weight of the cossettes in each step (30% overall water quantity), and (ii) the second consisted of four washing steps using only 5% water in each step (20% overall water quantity).

5 PULSED ELECTRIC FIELD ASSISTED PRESSING OF SUGAR BEET SLICES 61 The effect of varying the pressure has been studied for two values of the pressure: 5 and 10 bar Calculations and data analysis Following are the relationships used for calculating the different parameters. The juice yield Y may be given in terms of the mass of juices in kg, or in a normalised form Y *, by the following equations: Y ¼ X m juices (1a) Y ¼ 1 m P cake mjuices þ ¼ (1b) m coss m coss where: P m juices is the sum of juice masses in kg, before and after the PEF treatment, respectively; is the overall portions of juices lost in pipes and plates in kg; m coss is the initial mass of cossettes in kg; and m cake is the mass of the cake after all pressing steps. The composition of fresh cossettes, intermediate cake, and final pulp is customarily defined by the following relationships: M C ¼ 100 d W (2) S 0 S ¼ S SM C ð100 S S Þ N SS ¼ 100 ðm C þ S 0 SÞ (4) where: d W is the dry weight in kg per 100 kg of fresh cossettes in %; M C is the moisture content of the cossettes in %; S S is the soluble solids content in 1Brix, i.e. kg of soluble solids per 100 kg of juice; S 0 S is the soluble solids content in material, i.e. kg of soluble solids per 100 kg of material (cossettes, cake or pulp); N SS represents the solid non-soluble substances in kg per 100 kg of material. The initial moisture content of fresh cossettes is designated by M Co in %, and M Co is the normalised form of M Co ¼ M Co=100: The quality of a given juice is characterised by its purity in % and is given by the following formula: (3) The normalised form of this curve is given in Fig. 5. It should be noted that this curve does not take into account the amounts of juice losses that represent 5 7% of the initial weight of cossettes. This result illustrates the successful pilot scale-up of the process with respect to the laboratory-scale result obtained for the same pressing and treatment conditions (Bouzrara & Vorobiev, 2000). Table 2 summarises the scale-up performances based on the present pilot scale study, and shows that using 45kg (i.e times laboratory-scale) or 15 kg (i.e. 375 times) of initial cossettes, similar juice yields were obtained compared to laboratory scale yields from only 40 g of cossettes. One should bear in mind, however, that to confirm these results, the process must be scaled-up even further. In order to do so, continuous flow equipment, which allow complex pressing effects of a few hundreds of kilograms to a few tonnes of cossettes, are needed for future work. Masse of extracted juice, g PEF st press 2nd press 3rd press Time, min Fig. 4. Typical juice extraction kinetics during a 3-press process; pulsed electric field (PEF) applied at the end of the 1st pressing stage P ¼ 100 S C (5) S S which is the ratio of sugar content S C to the overall amount of soluble substances S S in the same juice. 3. Results and discussion 3.1. Performance of pulsed electric field-assisted pressing process Normalised overall yield st press 2nd press 3rd press PEF Kinetics of juice yield A typical acquisition curve describing the kinetics of juice yield during pilot scale pressing is given by Fig Time, min Fig. 5. Normalised accumulated mass of pressing

6 62 A.B. JEMAI; E. VOROBIEV Juice yield following pulsed electric field-assisted pressing at 5 bar To illustrate the type of performance that can be obtained in a PEF-assisted pressing process, consider the results from pressing at 5 bar given in Fig. 6. At this pressure, an average first pressing juice yield of about 29% can be achieved. Applying the PEF allowed an additional 48% (2nd and 3rd press) to be obtained. The characteristics of the different juices are summarised in Table 3. It can be seen that, compared to the 1st press, juices following the PEF treatment (2nd and 3rd press) systematically have higher purity values Table 2 Scale-up performance of pilot scale results compared to laboratory scale data Scaling factor 1st juice 2nd juice Yield, % Purity, % Yield, % Purity, % NA 495 NA y y Using 40 g of grated cossettes (Bouzrara, 2001). y Corresponding to 2nd+3rd juices with 3rd juice representing an average of 10% yield. Normalised juice yield Pressing Washing 1st juice 2nd juice 3rd juice 1st wash 2nd wash3rd wash Fig. 6. Juice yield during a typical 3-press/3-wash process at 5 bar (error bars are standard deviation values based on 13 experiments) (90 93% compared to 96 97%, respectively). The colour of juices following the PEF is systematically three to four times lighter than that of the juice before the PEF and factory juices. The colour was measured in ICUMSA units (International Commission for Uniform Methods of Sugar Analysis), which is based on spectrophotometer colour measurements, reflecting impurities and fine suspended particles contents in juices; this observation is confirmed by the colour difference of sugar crystals obtained from juices following PEF treatment compared to factory juice crystals (Table 4). This is why PEF juices are preferably handled separately in order to take advantage of this quality difference. Furthermore, maximising juice quantities after PEF should accordingly be considered. Other important criteria to consider in characterising the performance of the pressing process are the characteristics of the final pulp. For instance, significant amounts of non-sugars should have remained in the pulp following all pressing operations; this explains the high juice purity obtained following PEF treatment. Table 5 gives the residual pulp contents estimated based on experimental data and extrapolated to 2% sugar content of pulp. The general tendency is that, compared to industrial values, PEF pulp contained three to five times more a-amino nitrogen and two to three times more sodium and potassium (Anonymous, 2002). Table 4 Colour of factory and pulsed electric field (PEF) juices and final crystals in the international commission for uniform methods of sugar analysis (ICUMSA) Units Juice Colour, ICUMSA units Initial product before concentration Crystals Thick juice from factory y PEF juice Summarised from SUBEEP report (Anonymous, 2002). y Béghin-Say factory at Chevrière, France. Table 3 Characteristics of different juices for a 3-press/3-wash process at 5 bar Stage of treatment Juice yield, % Soluble solids, 1Brix Purity before purification, % Purity after purification, % 1st press, before PEF nd+3rd press, after PEF st wash nd wash rd wash PEF, pulsed electric field.

7 PULSED ELECTRIC FIELD ASSISTED PRESSING OF SUGAR BEET SLICES 63 Table 5 Estimated residual contents of pulp (per 100 g) following pulsed electric field (PEF) versus extraction process (extrapolated at sugar content S C ¼ 2%) Pulp 3.2. Optimising the process Residual content, g/100 g pulp Potassium Sodium a-amino nitrogen Sugar PEF pulp y Factory pulp z p2 Summarised from SUBEEP report (Anonymous, 2002). y Extrapolated values from experimental data. z Common industry values. From the present study, it appeared that following the three pressing steps, the pulp still contains significant amounts of juice and sugar; so pressing alone is not sufficient to obtain all the juice and sugar initially contained in the cossettes. Consequently, ways to optimise this process were explored. For example, one or more washing steps have been performed in order to extract further amounts of sugar remaining in the pulps. Optimising the level of treatment may also reveal the limits of the PEF treatment in achieving adequate juice yields. Finally, studying the impact of increasing the applied pressure on the juice yield and characteristics was explored Optimising the electric treatment The juice yield versus energy consumption curve, given by Fig. 7, shows that the PEF effect is significant up to an energy input of about 2 kj/kg; this effect levels off beyond 36 kj/kg. An optimum energy consumption range between these two limits is thus recommended for obtaining adequate juice yield with minimal consumption. It should be noted that with more appropriate pressing equipment (e.g. with complex shearing effects), this optimal range might further be reduced; in addition, the yield could be further enhanced owing to the complex pressing effects with some equipment such as belt or screw press Additional washing operations Due to the limits of the equipment used, three or four washing steps (batch) were needed to approach actual industry performances. From industry standards, wash is allowed using water representing 10 40% of the initial weight of cossettes (van-der-poel et al., 1998). Ideally, a single wash step using an amount of fresh water equivalent to 10% of the initial weight of the cossettes is to be performed leading Normalised juice yield Optical energy consumption zone Energy consumption, kj/kg Fig. 7. Juice yield (in normalised form) after the 3-press process as a function of energy input (optimisation experiments):, model;, experiment to extracting almost all sugar content of the pulp (o2% of pulp weight). This operation is best performed using continuous flow equipment. It is worth noting that to achieve actual industrial performance, it has been projected that at least 85% juice yield has to be obtained during the pressing steps in order to reduce the washing to only one step (see Section 3.3). (i) Three step washing operations Average results from a 3-press/3-wash process at 5 bar are given in Table 6; about 29% juice yield is achieved after the first pressing stage, while the second and third press juices (2nd press+3rd press) account for from 47% to 48% yield. Each subsequent wash step yields about 10% liquor. Despite washing and pressing limitations (i.e. simple pressing, short time washing, etc.), this illustrates that more than 10% of wash liquor could easily be extracted from the washed cake using more appropriate equipment. The data described here concern just the juice yield; one should further look at sugar yield and the purity aspect of the process. Characteristics of the different juices are shown in Table 6 and Fig. 8. The general tendency is that juices after PEF treatment have higher soluble solids content and purity. In fact, juices of the first pressing (before PEF treatment) have purity ranging from 90% to 93%, equivalent to that of factory juices; however, cold pressing juices are more concentrated in sugar: Brix compared to Brix for factory juices. On the other hand, press juices obtained following PEF treatment have systematically higher soluble solids content and purity (96 97%) before purification. As shown in Fig. 9, the difference in purity between 1st pressing juice and that of 2nd pressing is between 3% and 35%, while the difference between 3rd pressing juice and 1st wash liquor is about 25% and 2%, respectively.

8 64 A.B. JEMAI; E. VOROBIEV Table 6 Example of sugar loss reduction by one additional washing step using less wash water (three times 10% of initial weight of cossettes compared to four times 5%); process 1 versus process 2 corresponding to three and four washing steps, respectively Stage of treatment Process 1 : 3 press 3 wash Process 2 y : 3 press 4 wash Juice yield, % Sugar, % Soluble solids, 1Brix Purity, % Juice yield, % Sugar, % Soluble solids, 1Brix Purity, % 1st press, before PEF z nd+3rd press, after PEF st wash nd wash rd wash th wash Overall y y M C, moisture content; S C, sugar content; S S, soluble solids content; N SS, non-soluble solids. Initial composition (per 100 kg beets): MC ¼ 775%; S C ¼ 175%; S S ¼ 19 1Brix; N SS ¼ 432%: y Initial compositions (per 100 kg beets): M C ¼ 779%; S C ¼ 175%; S S ¼ Brix; N SS ¼ 377%: z PEF, pulsed electric field. y Sugar extraction (overall sugar yield divided by initial sugar content) of 9354% and 9703%, respectively. 98 Pressing Washing 96 Juice purity, % st juice 2nd juice 3rd juice 1st wash 2nd wash 3rd wash 4th wash Fig. 8. Purity of different juices for a typical 3-press/4-wash process at 5 bar (error bars are standard deviation values based on 43 experiments) Based on the present work, typical performances for a 3-press/3-wash procedure (at 5 bar) are summarised in Table 6; within the limits of the equipment used, this result demonstrates that a significant portion of sugar initially contained in the cossettes can be extracted at ambient temperature (about 935% sugar yield). (ii) Four step washing operations With the equipment at hand, optimising the washing procedure consisted of increasing the number of wash steps while using less water or using the same amount of new water while recycling the previous wash liquor. For example, 4-wash steps using only 5% of fresh water each step have been tested. Another scenario, consisting of 4- wash steps using 10% fresh water, which includes recycling of wash liquor from a previous run, has also been tested. Results from the optimisation of the number of wash steps are summarised by Table 6. It can be seen that sugar loss is reduced from 65% to 3% by increasing the number of wash steps while reducing the amount of wash water. Furthermore, recycling one wash liquor to increase the number of wash steps while using 10% of water may lead to a further decrease in losses. The major difference lies in the amount of water used in both cases (4 5% compared to 3 10% of the initial weight of cossettes).

9 PULSED ELECTRIC FIELD ASSISTED PRESSING OF SUGAR BEET SLICES Increasing the pressure One way to optimise the pressing process was to increase the pressure. As can be seen in Table 7, the overall juice yield did not seem to vary; however, the sugar yield slightly increased from 935% to 963%. With the equipment at hand, the applied pressure seems to effect the 1st press (i.e. from 29% to 335% for 5 bar up to 10 bar, respectively). In other words, more first pressing juice is obtained to the detriment of juice amount after PEF treatment. So in order to take advantage of this observation (specific to the equipment used), the duration of the first pressing stage should be decreased while increasing that of the second stage. The overall juice yield is not affected, but more juice having better quality is accumulated (i.e. more sugar can be extracted). Applying Purity difference, % nd juice 3rd juice 1st wash Fig. 9. Purity difference between the juices from the 1st press and that of the 2nd press, 3rd press, and 1st wash, respectively (error bars are standard deviation values based on 43 experiments) a higher pressure led to the final pressed pulp having higher dry weight and lower sugar loss (d W of 25% and sugar loss of 066 kg per 100 kg of cossettes compared to d W of 225% and sugar loss of 114 kg per 100 of cossettes, for 10 and 5 bar, respectively). Increasing the pressure allows to decrease the number of wash steps and leads to more wash liquor than the amount of wash water; so, with more appropriate pressing equipment significant quantities of wash liquors are obtained leading to a high value of d W for the pulp. This avoids the need for subsequent drying of the pulp Best possible scheme for pulsed electric field-assisted cold pressing In view of the results from the present work, a best possible scheme for a PEF-assisted cold pressing of sugar beet cossettes is given by Fig. 10. According to this flow diagram, an overall pressing juice yield of at least 85% is needed so that only one pressing step is required to ensure sugar extraction yield of at least 985%. This, in fact, corresponds to a loss of pulp of about 026 kg per 100 kg cossettes [i.e. equivalent to industry requirement (van-der-poel et al., 1998)]. Some interesting aspects of the process described by the flow diagram in Fig. 10 should be pointed out; for instance, significant quantities of high-quality juices are obtained following PEF application. Consequently, handling the different juices separately may be advantageous. Furthermore, intermediate pressed cossettes (i.e. following pressing stages) have a value for d W of about 40%. Attaining this high value is made possible by the Table 7 Effect of the pressure on the overall process performances; process 1 y versus process 2 z corresponding to 5 and 10 bar pressures, respectively Stage of treatment Pressure of 5 bar Pressure of 10 bar Juice yield, % Sugar, % Soluble solids, 1Brix Purity, % Juice yield, % Sugar, % Soluble solids, 1Brix Purity, % 1st press, before PEF nd+3rd press, after PEF 1st wash nd wash rd wash Overall y y M C, moisture content; S C, sugar content; S S, soluble solids content; N SS, non-soluble solids. PEF, pulsed electric field. y Initial composition: M C ¼ 775%; S C ¼ 175%; S S ¼ 19 1Brix; N SS ¼ 432%: z Initial compositions (per 100 kg beets): M C ¼ 777%; S C ¼ 177%; S S ¼ 192 1Brix; N SS ¼ 305%: y Sugar extraction (overall sugar yield divided by initial sugar content) of 9354% and 9630%, respectively.

10 66 A.B. JEMAI; E. VOROBIEV 100 kg of fresh cossettes : (Composition: M C = 77.8%; S = 18.85%; N SS = 3.35%; S S = 19.5 Brix); sugar = kg S PEF 1st juice : 35 kg S S = 18.5 Brix Sugar = 5.96 kg Purity = 92% Press I 2nd juice : 50 kg S S = 19.5 Brix Sugar = 9.46 kg Purity = 97% 15 kg of pressed cossettes: S C = 10.9% d W = 39.8% Washing 10 kg of fresh water Press II 1st juice 2nd juice Wash juice 97 kg of overall juice: S S = Brix Sugar = 16.8 kg (98.5%) Purity = 95% 12 kg of Wash juice: S S = Brix Sugar = 1.38 kg Purity = 95% 13 kg of exhausted pulp: S C = 1.96% d W = 34.8% Sugar = 0.26 kg (1.5%) Fig. 10. Best scheme scenario for a pulsed electric field (PEF)-assisted cold pressing of sugar beet cossettes: M C, moisture content; N SS, non-soluble solid content; M C, moisture content; S C, sugar content; S S, soluble solids content; d W, dry weight application of PEF during pressing with complex pressing equipment. In addition, despite the addition of wash water to the pressed cossettes, the final pulp also has a high value for d W (i.e. in the order of 35%). This eliminates the need to dry the pulp as customarily needed in industry (Jones, 1988). A correlation between the normalised juice yield Y * and the intermediate pressed cake dry weight d W has been established based on experimental data. It is given by the following equation: d W 100 ¼ 1 ðm Co ay Þ ð1 Y (6) Þ where: M Co is the normalised moisture content of fresh cossettes ( ); and a is a correlation constant ( ). The values of M Co and a were obtained using a fitting software (TableCurve 2D Windows V203, , by Jandel Scientific). As illustrated by Fig. 11, the semi-empirical model and experimental data fit well (coefficient of determination r 2 ¼ 0914). It can be seen that a 0% juice yield corresponds to initial raw cossettes, while at 85% juice yield, the value for d W is about 40%. Despite the limitations of the equipment used, a value for d W of around 34% has been attained; so, with more complex pressing effects equipment (i.e. shearing), it would be possible to attain 40% dry weight following the pressing steps. 4. Conclusion The present work conducted on a frame and platepressing equipment (45 15 kg of cossettes) validated laboratory scale results (40 g of cossettes) obtained using a filter-press cell. This successful scale-up confirmed that applying an intermediate pulsed electric field (PEF) treatment (following an initial pressing stage) significantly enhanced juice extraction yield from about 29% (1st press)

11 PULSED ELECTRIC FIELD ASSISTED PRESSING OF SUGAR BEET SLICES Dry weight of pressed cake d W / d W /100 = 1 ( Y *) / (1 Y*) r 2 = Normalised juice yield Y * Fig. 11. Correlation between juice yield and cake dry weight after a 3-press process:, model; determination, experiment; r 2, coefficient of to about 80% (overall yield per cent of initial weight of cossettes). It has been demonstrated that following the pressing operation at ambient temperature, sugar remaining in the pulp could be further extracted by performing a number of washing steps. Indeed, despite some equipment limitations and washing difficulties, it has been shown that losses can be significantly reduced to about 3% of initial sugar content. The juices extracted after PEF application (2nd and 3rd press) systematically have high purity (ranging from 95% to 97% before purification). While the purity of 1st press juice and that of 3rd wash liquor is equivalent to that of industrial juices, it is three to five points lower than PEF juices. The purity of the 1st step wash liquor is only one to two points lower than PEF juice. The purified overall juices obtained during the present process have ICUMSA units (International Commission for Uniform Methods of Sugar Analysis) three to four times lower compared to purified factory juices (e.g units compared to units). This colour difference has been confirmed following the crystallisation of PEF and factory juices to find that PEF crystals seem to be of bigger size and have lower ICUMSA colour (100 compared to 280 and 470 for factory juices). In addition to better juice quality, analysis of pressed and washed pulps concluded that after PEF treatment, including appropriate washing, substantially higher amounts of potassium, sodium and a-amino nitrogen remain in the pulp (i.e. extrapolation values at 2% sucrose content of pulp). In other words, less of these components are transferred to the pressed juice. From the results of the present work, further exploration is needed before any industrial implementation of the new process. Decreasing sugar losses by using more appropriate pressing equipment, which combines complex shearing effects and continuous juice production, should be considered. Further optimisation of the PEF treatment should be explored and fitted into a technical economic study. Acknowledgement The authors wish to thank SUBEEP programme partners for their financial and technical support in conducting this work. References Anonymous (2002). Pulsed electric field assisted pressing of sugar beet cossettes: a process of cold juice extraction. SUBEEP Programme Report. Compie` gne University of Technology, Compie` gne, France, GC-TAI Bouzrara H (2001). Ame lioration du pressage de produits ve ge taux par Champ Electrique Pulse : cas de la betterave à sucre. [Enhancing pressing of vegetable products by pulsed electric fields: case of sugar beet.] PhD thesis, Universite de Technologie de Compie` gne, France Bouzrara H; Vorobiev E (2000). Beet juice extraction by pressing and pulsed electric fields. International Sugar Journal, 102(1216), Bouzrara H; Vorobiev E (2001). Non-thermal pressing and washing of fresh sugarbeet cossettes combined with a pulsed electrical field. Zuckerindustrie, 126(6),

12 68 A.B. JEMAI; E. VOROBIEV Eshtiaghi M N; Knorr D (2002). High electric field pulse pretreatment: potential for sugar beet processing. Journal of Food Engineering, 52, Jones G C (1988). Cossette pre-treatment and pressing. International Sugar Journal, 90(1077), van-der-poel P W; Schiweck H; Shwartz T (1998). Sugar Technology. KG, Verlag Dr. Albert Bartens, Berlin Vorobiev E; Andre A; Bouzrara H; Bazhal M (2000). Proce de d extraction de liquide d un mate riau cellulaire, et dispositifs de mise en oeuvre du dit proce de. [Process of cellular material liquid extraction and implementation setup of such a process] French Patent. No of 22/02/2000, France