Cassava Starch Prod Vietnam
Cassava Starch Prod Vietnam
Cassava Starch Prod Vietnam
DOI 10.1002/star.200800202
CIRAD, UMR 95 Qualisud, Montpellier, France b HUT, IBFT, Hanoi, Vietnam c ENSBANA-UB, GPMA, EA 4181, Dijon, France d CIAT, A.A. 6713, Cali, Colombia
1 Introduction
In Southeast Asia the rapid economic recovery ensured ideal conditions for strengthened agricultural commodity chains. It was particularly significant for cassava (Manihot esculenta Crantz) in Thailand [1] and later on in Vietnam [2]. This particular crop grown in many parts of Asia [3], is competitive with other starchy sources [4]. In Vietnam, the crop yield increased dramatically in the 1990s [5] and eventually reached the average of 15 t/ha in 2006 [6], particularly with the introduction of high yield varieties [7] for industrial purposes with new farming techniques [8]. Some
Correspondence: Guillaume Da, ENSBANA-GPMA, Universit de Bourgogne (UB), 01 Esplanade Erasme, Dijon, France, Phone: 133 3 80 39 66 99, e-mail: guillaumeda@gmail.com and Dominique Dufour, Centro Internacional de Agricultura Tropical (CIAT), Apdo Areo 6713, Cali, Colombia. E-mail: D.Dufour@CGIAR. ORG.
authors reported that nationwide the share of commodity production utilized for starch extraction was 17% in 1991 [9], 24% in 1998 [10], and up to 40-70% in 2005 with the largest portion produced in the 36 cassava processing factories [2]. The Vietnamese Ministry of Agriculture and Rural Development reported in 2005 that small-scale processing accounted for over 70% of the total units, with a starch processing capacity of less than 1 t of starch at 12% moisture wet weight basis (wwb) per day, as noted by IFPRI [10]. It has been reported that this processing scale may last for one or two decades in Vietnam [11]. In the Red River Delta wet starch is produced within clusters of craft processing villages [12] where there is a clear division among starch processors [5]. Certain households and enterprises specialize in one or a group of activities, and together they have a relationship that is both competitive and innovative [13]. Despite space limitations for processing activities, this model of production not only continues to grow locally but also expands to other provinces [5]. www.starch-journal.com
Research Paper
359
From a technological perspective, examining processes at both large or small scales indicates that there is diversity in cassava starch manufacturing [14]. The processing yield (kg of recovered dry starch divided by 100 kg of fresh roots), consumption of water and possible electrical requirements have been reported in previous diagnoses [15]. The processing yield range from 17% in Ivory Coast and in Colombia [16], 21% in Brazil [17] and up to 25% in Thailand [1]. Water consumption per kilogram of starch was reported to be in the range 21-40 L in Brazil [18], 30 L in India [19], or up to 50 L in Colombia [20, 21]. Electrical energy requirements per tonne of starch ranged from 14 kWh and 21 kWh in small- and medium-scale units in India [22]. Despite these figures revealing potential differences between processes, they remain difficult to compare because of the use of different methodologies to estimate their components. This study proposes a diagnosis where a range of selective measurements are applied to cassava wet starch processing units in Northern Vietnam. The objectives consist of counting and characterizing the processing units within a cluster of three processing communes, and by taking into account the following parameters: number of processing days per year, material balance, starch yield, water and electrical consumption, labor, capital investment, equipment depreciation, and production costs. The objective is to not only assess the different technologies used locally, but to develop a method of comparison that can be expanded to other starch manufacturing processes in other contexts.
Fig. 1. Flow chart of cassava wet starch processing at a small scale in Vietnam in 2006.
has the potential for market linkages [24]. Furthermore, this province has the largest production of cassava wet starch in the Red River Delta [5] and the selected communes have extensive experience in trading and processing cassava and cassava-based products [24].
360
G. Da et al.
Fig. 2. Cylindrical rasper (type A and B) for cassava starch processing at a small scale in Vietnam.
Fig. 3. Vertical stirring tank reactor (type B) for cassava starch processing at a small scale in Vietnam.
Fig. 4. Rasper/extractor (type C) for cassava starch processing at a small scale in Vietnam.
www.starch-journal.com
361
wood and serrated with fine wires (3 mm in height). A pipe inlet was placed on the upper edge of the rasping chamber to spray water onto the peeled roots. Extraction: The extractor for system A consisted of a vertical circular basket (0.1 m3) where the pulp was stirred by hand with water prior to being sieved through a cotton filter cloth (aperture of 70 mesh-210 mm). The extractor for system B [25] consisted of a vertical stirring axle placed in an aluminum tank (0.8 m3) flanked by two baffles which were fixed 10 cm from the bottom of the tank (Fig. 3). A cotton filter cloth (with an aperture of 70 mesh) was placed on the bottom screen of the extractor. The extractor for system C (Fig. 4) consisted of a vertical stirring-filtering reactor (0.1 m3 in volume). The movement of the vertical shaft, which held in place the rasping disc with the two paddles used for extraction, allowed the rasping and extraction stages to work simultaneously [25]. Extractor C was equipped with a cotton filter cloth (aperture of 70 mesh) and a 10 mesh (2 mm) screen. For systems B and C, intermediate tanks lined with ceramic tiles were used between the separation and the settling stage in order to ensure a decantation of starch milk [1]. Settling and dewatering: For the three systems, two consecutive sieves (aperture of 70 and 10 mesh for the upper and lower layers respectively) fixed on wooden frames were used before sedimentation in settling tanks. The absorption stage occurred after collecting the yellowish green tint layer [26] (locally called black starch) on top of the wet starch after settling stage. This fresh black starch was discarded from the tank, and weighed. Samples were collected in triplicate and dried for analysis. Usually, fresh black starch is transferred subsequently by processors into appropriate concrete tanks for a second settling stage, where a natural fermentation occurs, with a reduction in pH from 6 to 4. Finally, the resulting product (fermented black starch) is collected and boiled before being incorporated in the daily food ration for pigs. The viscosities of both fresh and fermented black starches were not studied. Sedimented wet starch was covered with a layer of cloth and dried coal residue (up to 12 kg/m3) in order to absorb sufficient amounts of water before being cut into blocks to be stacked up moist (45% moisture content, wwb) onto clay bricks. Water pumping: Unfiltered water was supplied for the washing stage and was directly drawn from wells; whereas the extraction stage required filtered water previously drawn and poured in a concrete tank filled with sand. For the entire process, three types of pumps were used: One (Qmax = 70 L/min, P=0.75 kW) to draw under-
ground water (types A-B-C); one (Qmax = 260 L/min, P=0.75 kW) to pump filtered water up to the extractor (types B-C) and one (Qmax = 100 L/min, P=0.37 kW) to pump starch milk from the intermediate tank into the main settling tank (types B-C).
2.2 Methods 2.2.1 Mass balance of the wet starch manufacturing process
Measurements were conducted for particular stages of the manufacturing process (Fig. 5) depending on access and relevance to samples throughout this process [27]. The quantity parameters (weight, time) as well as quality parameters (analysis) were then reported in detailed flow sheets in order to compare the mass balances between the different processing types. Organizing the trials and moisture content: Diagnostic trials were repeated in triplicate with the same households during the 2006 season. For each trial, 15 t of high yield cassava varieties [2] were purchased and delivered by truck. The roots were then divided within the A, B and C processing units in order to be crushed simultaneously (Fig. 5). Up to 4 kg of roots were collected from the truck delivery, and then the roots were immediately cut into small pieces, mixed and dried at 607C for 48 h [28] for further analysis. Samples in triplicate which needed to be analyzed were also collected from the manufacturing process (Fig. 5) during the 2006 harvesting season and they underwent the same procedure. The moisture content of the products which did not need further analysis, was determined by drying 10 g of sample at 1057C for 24 h [29]. The amount of total solid waste carried by wastewater was measured in waste-water samples (500 mL) collected at regular intervals from the main sedimentation tank. The samples of waste-water were evaporated (without boiling), and dried at 1057C until a constant weight was reached. The difference between the weight of the empty container and the weight of the container after drying represented the total solids carried in the waste water. Water consumption measurement: The water consumption was measured (Fig. 5) by water meters previously placed on the water inlets of the washers and a type B or C extractor. For manual extraction (type A), a derivative device system was set-up with one centrifugal pump (Qmax = 100 L/min, P=0.37 kW) and one watermeter. The volumes of suspensions (starch milk, wastewater) in the sedimentation tanks were also reported (Fig. 5). www.starch-journal.com
362
G. Da et al.
Fig. 5. Flow chart diagram of wet starch production at a small scale in Northern Vietnam. The weight symbol corresponds to the solid materials which have been weighed and sampled. The arrow symbol marked water corresponds to a water inlet where water meters were set-up. The ruler symbol corresponds to the ruler used to measure the volume of liquid in the settling tanks.
2.2.2 Analysis
Determination of starch content: The starch content of the solid samples (Fig. 5) was measured using an enzymatic colorimetric method [30]. The results are given in% of starch per kilogram of dry matter. The starch content in the starch milk suspensions was obtained by a density method. Two hydrometers were used (Dujardin-Salleron, Arcueil, France, model 10001030 6 0.2 g 10-3 m3 and model 1000-1100 6 0.5 g 10-3 m3) to measure the density (at 207C) of aliquots (0.500 L) collected at regular intervals from the outlets of type B and type C extractors throughout the duration of the extraction process. Crude fiber content: The fiber content was determined from the loss of ignition of the dried residue that remained after the digestion of cassava flour (2 g) with 1.25% aqueous H2SO4 and 1.25% aqueous NaOH [31]. The results are given in% of crude fibers per kilogram of dry matter. Ash content: The ash content was calculated after 1g of sample was heated at 5507C for 3 h [32]. The results are given in% of ash per kg of dry matter.
Scanning electron microscopy (SEM): Dehydrated samples (starch granules, cassava bagasse, black starch) were sprinkled on double-sided sticky tape, mounted on circular aluminum stubs, coated with 35 nm of gold-aluminum, and then photographed in a scanning electron microscope (JSM 820 Jeol, Tokyo, Japan) at an accelerating voltage of 20 kV. The granule size was measured.
Particle size analysis: The sizes of starch granules and particles in wet starch and black starch were determined by laser diffraction analysis (Mastersizer 2000, Malvern Instruments LtD, Worcestershire, UK) with a refractive index of 1.53. The particle size is expressed in mean volume in micrometers [33].
Electrical consumptions measurements: The study focused on the rasping and extraction stages which both required three-phase electric power. This power was measured by a current analyzer (model CA 8230, Chauvin Arnoux, Paris, France) during the trials conducted in 2006. The device was placed on the three phases electrical outlet in order to record the real power (every 5 s) of the rasper and the extractor while it was working. www.starch-journal.com
363
sample means were determined using the Students t-test or ANOVA test for multiple comparisons at a 95% confidence level. The statistical analyses were performed with Statistica v7.1 software (Statsoft, Inc., Maisons-Alfort, France).
3 Results
3.1 Production characteristics 3.1.1 Variability of the quality of raw material
The quality of the fresh roots for processing starch depended on the dry matter content of the roots (Fig. 6). This resulted in different levels of starch content, which is known to be closely correlated not only to the apparent density [34, 35] but also to the dry matter content of the roots [28, 36]. During the trials, the roots had a dry matter content, crude fiber content and starch content in the range of 38.9-44.3%, 2.33.9%, and 80.6-84.7%, respectively. The lowest values in dry matter and starch corresponded with the end of the processing season, when the rainy season started (April).
Fig. 6. Distribution of the dry matter content of the cassava fresh roots used for wet starch processing at a small scale in Vietnam (2006-2007).
www.starch-journal.com
364
G. Da et al.
Starch/Strke 60 (2008) 358372 Tab. 2. Balance sheet for the cassava starch manufacturing process at a small scale in Vietnam in 2006 (kg of dry weight). Material Fresh roots Peeled roots Peels, dirt Cassava bagasse Black starch fraction Wet starch Total solids carried in waste-water Type A [kg] Type B [kg] Type C [kg]
Tab. 1. Distribution of the three types of technologies within the surveyed communes in 2005 in Vietnam. Commune Number of households producing wet starch 514 184 35 Distribution of households types A [%] 2 58 0 B [%] 75 36 71 C [%] 23 6 29
100 100 100 98.4 6 1.1a 97.6 6 1.8a 97.8 6 1.7a 1.6 6 0.4a 1.5 6 0.4a 1.5 6 0.7a a a 21.4 6 1.6 20.1 6 0.1 13.8 6 1.0b a a 3.1 6 0.8 3.6 6 1.5 3.3 6 1.1a 63.5 6 2.9a 62.2 6 3.1a 64.8 6 0.8a 10.4 6 1.3a 12.6 6 2.1a 16.5 6 1.9b
switch from type A to type B and subsequently to type C processing types. This evolution has been mainly motivated by increasing the producing capacities and reducing the labor input [23].
The data provides the average of three trials conducted from January to March 2006 (processing season). The standard deviations are indicated with a 6 sign. Within each line, statistically significant differences at a = 5% level are indicated with superscript letters (e.g., a, b).
3.2 Mass balance of wet starch 3.2.1 Balance sheet of wet starch
The calculation of the mass balance (Tab. 2) resulted in different coefficients of variation. Peels and black starch (in the range 22-46%) demonstrated that they have been affected by the fluctuation of the quantity of dirt on the roots. Consequently, the calculation of the mass balance can be achieved by weighing only peeled roots (or cassava pulp), cassava bagasse and wet starch, which all corresponded to low variation coefficients (1.8%, 7.4% and 4.9%, respectively). Furthermore, the mass balance of type C (Tab. 2) revealed that the greater quantity of dry matter liberated during the extraction stage was followed by a greater amount of total solids carried in the waste-water compared with the A and B types. Tab. 3. Yield components for three types of cassava manufacturing processes from Vietnam at a small scale in 2006. Yield components Processing yield Overall starch recovery Rasping effect Starch extraction efficiency A [%] 25.4 6 1.5a 74.4 6 3.9a 85.1 6 0.3a 86.3 6 1.5a B [%] 25.0 6 1.1a 73.3 6 4.0a 85.0 6 2.2a 87.3 6 3.9a C [%] 27.0 6 1.7a 76.1 6 1.9a 93.1 6 0.9b 92.6 6 0.7b
The data provides the average of three trials conducted from January to March 2006 (processing season). The standard deviations are indicated with a 6 sign. Within each line, statistically significant differences at a = 5% level are indicated with superscript letters (e.g., a, b).
365
Tab. 4. Water-consumption for types A, B and C at a small-scale in Vietnam in 2006 (L/kg of starch at 12% moisture content, wwb). Processing stage A [L/kg] 2.0 6 0.9(****) 0.0 6 0.0(***) 14.0 6 0.4c(*) 16.0 6 0.4c B (no recirculation) [L/kg] 2.0 6 0.9(****) 0.0 6 0.0(***) 11.1 6 0.6b(**) 13.1 6 0.6b B (recirculation) [L/kg] 2.0 6 0.9(****) 0.0 6 0.0(***) 7.9 6 0.3a(**) 10.0 6 0.3a C [L/kg] 2.0 6 0.9(****) 18.8 6 0.4d(**) 20.8 6 0.4d(**)
Average and standard deviation obtained from 4 trials (*), 6 trials (**), 8 trials (***), 11 trials (****) conducted from January to December 2006. The standard deviations are indicated with a 6 sign. Within each line, statistically significant differences at a=5% level are indicated with superscript letters (e.g., a, b, c, d). The data which is similar along the same line corresponds to the same measurement. The data reported for type C during the extraction stage includes the volume of water used for both rasping and extraction stages which worked simultaneously.
water for both the rasping and extraction stages (Fig. 4), where 6.1 ( 6 0.1) and 12.7 ( 6 0.1) L/kg starch were supplied through the upper water inlet (simultaneous rasping-extraction phase) and through the lower inlet (extraction phase without rasping).
different components indicated that the determination of the composition was not sufficient to characterize the differences between A-B-C starches, like previously noted elsewhere [1, 28]. The mean volume of the starch granules in wet starch was 24.15 mm (62.81).
366
G. Da et al.
Fig. 7. Up: Starch granules in cassava pulp before extraction (left) and in cassava bagasse after extraction (right). Down: Starch granules (left) and black starch fraction (right). Tab. 5. Performances of raspers and extractors in Vietnam. Parameters Type of rasper Drum Disc Rotation per minute [rpm] Linear speed on the edge [m/s] Used linear speed [m/s] Rasping surface [m2] Sieving surface [m2] Surface6contact time [m2] 4,400 2,400 50.7 72.8 Type of extractor B 140 8.6 C 2,400 57.8 The drop in the extraction capacities of both systems A and B (Tab. 6) depended mainly on the technologies used. The type C extraction did not require loading and its higher capacity, which almost equaled the capacity of the previous stages (Tab. 6), was characterized by a greater velocity of the blades (Tab. 5) as compared with the type B extractor. The settling and dewatering capacities were similar for the three systems A, B and C in a range of 6191 kg/h of fresh roots. No significant innovations have been reported locally to speed up the starch settling during sedimentation.
6.9
49.9
1.24 5.8
0.22 17.6
(1,039 6 39 kg/h) were not significantly different and both technologies were limited by the manual loading time, which equaled the time to load the roots into the washer.
367
Fig. 8. Starch milk density profile during type B extraction (left) and type C extraction (right) from Vietnam. The symbol corresponds to the experimental point (average of triplicates samples with repeated trials) with standard deviation.
Tab. 6. Capacity (kg of entering material per hour) of washing, rasping and extraction stages during the cassava wet starch manufacturing process in 2006 in Vietnam. Processing stage A [kg/h] Washing Rasping Extraction Processing types B [kg/h] C [kg/h]
Tab. 7. Comparison of electrical requirements between three types of cassava processing systems for producing 1 t of commercialized Starch (at 12% moisture content, wwb) in Vietnam in 2006. Parameters Cylindrical rasper (types A-B) Front sieving extractor (type B) Rasper extractor (type C) 15 13.0 (64.6) 55.3
1,140 6 106 1,140 6 106 1,140 6 106 1,053 6 192c 1,053 6 192c 273 6 34a 410 6 46b 864 6 83c
Engine power 15 3 [kW] Real power consumption 10.4 (62.0) 2.3 (60.4) [kW] Electrical energy 19.8 10.6 [kWh/t]
The data provides the average of three trials conducted from January to March 2006 (processing season). The standard deviations are indicated with a 6 sign. Statistically significant differences at a=5% level are indicated with superscript letters (e.g., a, b, c). The data which is similar along the same line corresponds to the same measurement. The data reported here for type C at extraction stage includes the capacity obtained for both rasping and extraction stages which worked simultaneously. The slurry is poured in an intermediate tank before being used at the beginning of the next batch. This corresponds to a 29% (63) reduction of the total water used compared to normal extraction with the same reactor.
The standard deviations are indicated with a 6 sign. rasping surface to increase the rasping efficiency and rasping capacities (up to 1,789 kg/h). This resulted in a heterogeneous demand of power (Tab. 7) with a maximum of 16.2 kW, which made the engine capacity suitable for this stage. By using the rasping disc, the operator limited manual hazard risks. High variation coefficients in power consumption were obtained (Tab. 7) where the highest demand in power (25 kW maximum) occurred during the rasping stage. The minimum power (3.9 kW) corresponded to the time (5 s) when the operator emptied the cassava bagasse after extraction. A high demand in the number of three-phases of the electrical units (Tab. 7) could have resulted from a greater velocity of the rasping disc (Tab. 5) rather than the rasping cylinder. Thus, considering an engine yield of 75% to convert electrical energy into mechanical energy, the measurement of the www.starch-journal.com
368
G. Da et al.
Starch/Strke 60 (2008) 358372 Tab. 8. Labor input for 1 t of starch (at 12% moisture content, wwb) production in small-scale cassava starch industries in Vietnam in 2006 (man-hours). Unit operations Processing types Type A [Mh]a Reception of fresh roots Washing the fresh roots Sorting and Transferring the peeled roots to rasping Rasping Transporting the pulp Extraction Settling and drain off Dewatering and cutting Total Mha = man-hours. 3.8 4.6 3.2 3.6 5.3 22.0 12.9 3.7 58.7 Type B [Mh]a 3.8 4.6 3.2 3.6 5.3 6.4 12.9 3.7 43.8 Type C [Mh]a 3.8 4.6 3.2 1.6 0.0 2.5 6.9 3.0 23.2
type C equipment revealed that 40% of the mechanical energy was used for rasping and 60% was used for extraction. This suggests that both types of raspers required similar levels of power; however, a higher level of electrical power was required for type C extraction as compared to type B (Tab. 7). The type B extractor required a maximum real power of 3.4 kW at the end of the process, when the fibrous residues became dryer. The capacity of the engine of 3 kW was efficient and the average mechanical power required for this stage was 1.6 kW.
cause each processing unit usually operates within the same household. So, starch manufacturing at small scale has not only been profitable (Tab. 9), but also represents a local employment opportunity [40].
4 Discussion
The future of producing starch at a small scale was limited in Vietnam in 1999 [41]; however, the capacity to expand this activity has been revealed in this study and can be achieved by substituting capital investments for labor reduction. From a technological perspective, raspingextraction (type C) resulted in increasing process efficiency with low quantities of starch contained in cassava bagasse compared to other extraction technologies locally or elsewhere [42]. This could be attributed to better rasping efficiency, followed by better extraction efficiency, in order to liberate the starch granules trapped in the network of fibers. Furthermore, greater capital-intensive production methods continue arising such as new mobile continuous washers, which can be fed from the bulk delivery and then directly connected to the raspingextracting machine [25]. Despite making investments for the new washing technology which may result in the continued optimization of the capital and labor for production [43], it may require a greater source of electricity, which can be a constraint to extend these technologies in the future either to other rural areas in Vietnam or eventually to other developing countries. The processors, however, endure remaining issues to maximize starch production. Firstly, the sedimentation stage is still a bottleneck for the manufacturing process. www.starch-journal.com
369
Tab. 9. Cost structure, profit and income generation of small-scale units for one month in Vietnam (2006). Items Unit Type A Quantity Starch By-product Total returns Raw material Electricity Contengencies Fixed costs Laborers Production costs Profit B/C ratio Income generation [t] [t] [month] [t] [month] [month] [month] [Mma] [month] [month] [month] [month] 10 3 1 33 0 1 1 1 1 1 1 1 Value [US$] 1,714 43 1,757 1,431 16 31 51 92 1,621 136 1.08 228 Type B Quantity 9 3 1 32 0 1 1 1 1 1 1 1 Value [US$] 1,657 42 1,699 1,383 24 29 60 66 1,562 137 1.09 202 Type C Quantity 19 5 1 66 0 1 1 1 1 1 1 1 Value [US$] 3,381 85 3,467 2,795 90 59 96 79 3,119 348 1.11 427
1 2 3 4 5 6 7 8 9
Profit equals to total returns (1) minus production costs (6). B/C ratio (8) corresponds to total returns (1) divided by the subtraction of (1) and (7). Income generation (9) is the addition of (5) and (7). Mma = man-months. Starch is at 12% moisture content wwb.
Some options such as concentrating the starch milk by using hydrocyclones [44, 45] or using the sour liquid method similar to the sweet potato starch production in China [46, 47], could be investigated locally to reduce the settling time [48, 49]. Secondly, water consumption remains critical for the manufacturing process. This study shows promising results on the reduction of water consumption by re-circulating starch milk during extraction with Type B. Although it resulted in a reduction in the quantity of water compared to the other types (A, C) this reduction remains lower than the 50% announced by processors who use hydrocyclones at similar scales [44, 45]. Thus, the possibility of re-circulating starch milk with system C or developing a continuous rasping-extractor might be additional opportunities to investigate. Thirdly, each household usually digs an individual well to pump ground water up to 80 m, and the waste-water (converted from the water consumed in the extraction stage) is then discharged from the settling tanks without treatment, as previously reported in Colombia [50, 51] or in India for sago starch [45]. The potential of wastewater as a nutrient-rich irrigation source for rice was evaluated at a small-scale in Vietnam [52], however, it has not been applied locally. Thus, this study shows greater production capacities with type C processing as a result of greater amounts of total solids carried in the waste-water. Consequently, particular attention must be paid on the water treatment by upgrading the technologies to type C. Households have limited space and therefore using bioponds may not be a practical option to advocate for wastewater treatment. [53]. Consequently, the treatment of waste-water may be investigated locally through more
economical solutions like anaerobic horizontal flow filters resulting in biogas production, such as in the case of Columbia where promising results are evident [54]. Finally, wet starch is commonly used directly for specific local demands [52, 55], and eventually it is stored under anaerobic conditions where its quality is subject to change [56, 57] before being reprocessed locally (after one year). With the new development of large scale units [5, 58], small-scale processors face a new challenge to reach quality standards required by end-users unless wet starch production becomes an intermediate scale product to reduce costs for further large-scale processing (purifying, drying, packaging) as reported previously in China [59]. It is important to note the distance between harvesting to processing (up to 200 km), which results in numerous transaction costs [5], and restrict potential price reduction for raw materials. In the production cost, these account for higher shares compared to other locations, either at a small scale [60] or at a large scale [1]. Moreover, from 2006, the high demand for cassava chips in China or starchy sources for bioethanol production result in an insufficient raw material supply for all starch processors [61] who have to handle a great variability of the quality of the roots.
5 Conclusion
The methodology of diagnosis proposed in this study applied to cassava wet starch, allowed the assessment of www.starch-journal.com
370
G. Da et al.
manufacturing units at a small scale in Vietnam. The results revealed that this production scale showed higher efficiency than similar scales in other locations. The main differences between the three processing types, which differ in the rasping and extraction stages, were in capacities, water consumptions, electrical requirements and capital-labor costs, with a tendency in adopting continuous processes for washing and rasping stages. Further capacity expansion for processors might be allowed by developing continuous rasping-extraction stages with higher concentrations of starch milk in order to overcome high local constraints on space limitation. A complementary study on socio-economic aspects (valuechain mapping, customers need on starch quality, access to credit) would help to better understand the organization of the production of wet starch. Finally, the methodology of diagnosis presented in this study can be used as a basis for other foodstuffs produced locally (canna or kudzu starch) or elsewhere (sweet potato starch) to supplement information for rural development issues.
[6] [7]
[8]
[9]
[10]
[11] [12]
Acknowledgements
The work reported in this study was supported by the Centre de Coopration Internationale en Recherche Agronomique pour le Dveloppement (CIRAD) as well as Hanoi University of Technology (HUT) and the University of Burgundy (ENSBANA-UB). We also acknowledge Dr. Thierry Tran (Cassava and Starch Technology Research Unit, CSTRU, Kasetsart University, Thailand) and Tereza Sanchez (Centro Internacional de Agricultura Tropical, CIAT, Colombia) for analysis. The authors are thankful for Dr. Dai Peters (CIAT), Loan Trinh Thi Phuong and Nguyen Khac Quynh (Vietnamese Agricultural Academy of Science, VAAS) for their help on craft villages.
[13]
[14]
[15]
[16]
[17]
References
[1] K. Sriroth, K. Piyachomwan, S. Wanlapatit, C. Oates: Cassava starch technology: The Thai experience. Starch/Strke 2000, 52, 439449. [2] K. Hoang, V. B. Pham, R. Howeler, J. J. Wang, N. N. Tran, K. Kawano, H. Ceballos: The history and recent developments of the cassava sector in Vietnam: Innovative technologies for commercialization. Concise papers of the Second International Symposium on Sweetpotato and Cassava, Kuala Lumpur, Malaysia, June 14-17. 2005, 2627, 2005. [3] S. N. Moorthy: Tropical sources of starch, in Starch in Food: Structure, Function and Applications (Ed. A.C. Eliasson) Woodhead Publishing, 321359, 2004. [4] K. O. Fuglie: Challenging Bennets law: the new economics of starchy staples in Asia. Food Policy 2004, 29, 187202. [5] T. A. Dao, D. T. Le, T. B. Vu, D. H. Dao: Pathways out of Poverty through Secondary Crops and Private Sector Processing
[18]
[19]
[20]
[21]
www.starch-journal.com
371
[22]
[23]
[24] [25]
[26]
[27]
[28]
[29]
[30]
[34]
[35]
[36]
en Recherche Agronomique pour le Dveloppement (CIRAD), Montpellier, France, 62, 2001. J. T. Sheriff, S. K. Nanda, M. S. Sajeev: Current status of cassava processing industries in south India, Technical bulletin n7 42, CTCRI. Thiruvananthapuram, Kerala, India, Central Tuber Crops Research Institute (CTCRI), 78, 2005. G. Da, Le Thanh Mai, D. Dufour, P. Marechal, A. N. K. Quynh: Participatory Assessment of Cassava Starch Processing at Small-Scale in Vietnam (Hoai Duc district, Ha Tay Province). Proceeding of the Regional Symposium on Chemical Engineering New Trends in Technology towards Sustainable Development held in Hanoi. Nov 30 - Dec 02. 2005. Hanoi, Vietnam: Science and Technics Publishing House, 4551, 2005. T. T. Dao: Rising from tradition: Agricultural clusters. Phat Trien Nong Thon - Rural development 2004, 6, 47, 3. C. Marouz, G. Da, D. Dufour: Inventaire des quipements de petite et moyenne capacits de production damidon de manioc par la voie humide. Proceedings from the 1st International Workshop held in Abidjian, Ivory Coast, June 4-7, 2007, inPotentialits la transformation du manioc (Manihot esculenta Crantz) en Afrique de lOuest, Actes de lAtelier international. (Ed. G. Amani, C. Nindjin, B. NZu, A. Tschannen, D. Aka), 127132, 2007. M. Grace, R: Cassava Processing. Agricultural Services Bulletin No.8, Plant Production and Protection, Food and Agriculture Organization of the United Nations (FAO), Rome, Italy, 155, 1977. C. Marouz, G. Da: Outil de caractrisation dune ligne de transformation alimentaire, cas de la production damidon de manioc. Proceedings from the 1st International Workshop held in Abidjian, Ivory Coast, June 4-7, 2007, in Potentialits la transformation du manioc (Manihot esculenta Crantz) en Afrique de lOuest, Actes de lAtelier international. (Ed. G. Amani, C. Nindjin, B. NZu, A. Tschannen, D. Aka), 133139, 2007. A. Dufour: Evaluation de la maturit du manioc daltitude pour la production damidon en Colombie. Thesis, University of Picardy, France (available at CIAT, Cali, Colombia), 2007. M. Le Thanh, H. Nguyen Thi, T. Pham Thu, H. Nguyen Thanh, L. C. Le Thi: Cac phuong phap phan tich nganh cong nghe len men (Analytical methods currently used in food fermentation), in Vietnamese. Science and Technics Publishing house, Hanoi, Vietnam, 331, 2005. I. L. Batey: Starch analysis using thermostable R-amylases. Starch/Strke 1986, 34, 125128. Modified by Mestres as reported in C. Mestres: Glification damidons de mas modifis thermiquement. Application la fabrication de ptes alimentaires sans gluten. Ph.D. Thesis. France, 1982. AOAC: Fiber (Crude) in Animal Feed and Pet Food; AOAC 962.09, 16th ed: Washington, DC, USA. 1, 1996. AOAC: Ash of Flour: AOAC 923.03, 16th ed: Washington, DC, USA. 1, 1996. H. D. Brittain: Particle-Size Distribution, Part I. Representations of Particle Shape, Size, and Distribution, P. Physics, Advanstar: Milford, NJ 08848, USA. 6, 2001. Z. Bainbridge, K. Tomlins, K. Wellings, A. Westby: Methods for Assessing Quality Characteristics of Non-grain Starch Staples (Part.2 Field methods). Natural Resources Institute, University of Greenwich. 34, 1996. L. Neamfug, K. Piyachomkwan, K. Sriroth: Simple method in estimating starch content of fresh cassava and sweet potato tuber, T.P.N. 009/1999). Cassava and Starch Research Technology Research Unit, Bangkok, 1999. A. Fernandez, J. Wenham, D. Dufour, C. C. Wheatley: The influence of Variety and Processing on the Physicochemical
and Functional Properties of Cassava Starch and Flour, in Cassava Flour and Starch; Progress in research and Development. (Ed. D. Dufour) CIRAD & CIAT, 263269, 1996. [37] F. Alarcon, D. Dufour: eds. Almidon agrio de Yuca en Columbia. Tomo 1: produccion y recomendaciones. Vol. 1, Centro Internacional de Agricultura Tropical (CIAT), Centre de coopration internationale en recherche agronomique pour le dveloppement (CIRAD); Cali, Colombia, 35, 1998. [38] A. Westby: Cassava utilization, Storage and small-scale processing, in Cassava: biology, production and utilization. (Ed. R.J. Hillocks, J.M. Thresh, A. Bellotti) CABI, Cali, Colombia, 281300, 2002. [39] J. O. Olaoye: Simulation of cassava grating systems for cottage industries in the middle belt of Nigeria. Proceeding of the 7th International Conference on Computers in Agriculture held in Orlando, Florida, USA. October 26-30, 1998, 991999, 1998. [40] V. Gottret, G. Henry, D. Dufour: Adoption et impact de la transformation du manioc en amidon aigre. Les cahiers de la recherche dveloppement 1997, 44, 3859. [41] F. Goletti, K. Rich, C. Wheatley: The cassava starch industry in Vietnam: Can small firms survive and prosper? The International Food and Agribusiness Management Review 1999, 2, 3, 345357. [42] A. Pandey, C. R. Soccol, P. Nigam, V. T. Soccol, L. P. S. Vandenberghe, R. Mohan: Biotechnological potential of agro-industrial residues. II: cassava bagasse. Bioresource Technology 2000, 74 (1), 8187. [43] A. Graffham, A. Westby: Cassava Starch. Case study: Benin, Ghana and Malawi, in Global Cassava Market Study: Business Opportunities for the Uses of Cassava. (Ed. dTp Studies Inc) Guelph, Ontario, Canada, 3975, 1999. [44] A. Manickavasagan, K. Thangavel: Single hydrocyclone for cassava starch milk. Agric. Mech. Asia Afr. Lat. Am. 2006, 37, 1, 8084. [45] R. C. Marder, D. S. Trim: Water conservation and effluent reduction in the cassava starch extraction industries using hydrocyclone technology, in Cassava, Starch and Starch Derivatives. Proceedings of the International Symposium held in Nanning, Guangxi, China. Nov 11-15, 1996. (Ed. R. Howeler, C. Oates, G.M. OBrien) Centro Internacional de Agricultura Tropical (CIAT), Cali, Colombia, 154168, 2000. [46] K. Piyachomkwan, K. Sriroth, K. Chinsamran, K. Laohaphattanalert, C. Oates: Development of a standard protocol for the processing of high quality sweetpotato starch for noodle making: Sweetpotato post-harvest research and development in China. Proceedings of an International Workshop held in Chengdu, Sichuan, PR China, Nov. 78. 2001: International Potato Center (CIP), Bogor, Indonesia., 160, 2004. [47] W. H. Timmins, A. D. Marter, A. Westby: Aspects of Sweet potato Processing in Sichuan Province. P.R. China, in Product Development for Root and Tuber Crops. Volume 1. Asia. Proceedings of an International Workshop, held in Viasayas State College of Agriculture (ViSCA); Baybay, Leyte, Philippines, April 22-May 1, 1991. (Ed. G. J. Schott, S. Wiersema, P. Ferguson, I) International Potato Center (CIP), Lima, Peru, 217228, 1992. [48] M. S. Sajeev, R. Kailappan, V. V. Sreenarayanan, K. Thangavel: PH-Postharvest technology: Kinetics of gravity settling of cassava starch in its aqueous suspension. Biosys. Eng. 2002, 83, 3, 327337. [49] M. S. Sajeev, S. N. Moorthy, R. Kailappan, V. Sunitha Rani: Gelatinisation characteristics of cassava starch settled in the presence of different chemicals. Starch/Strke 2003 55 (5), 247257.
www.starch-journal.com
372
G. Da et al.
[50] V. Gottret, G. Henry, D. Dufour: Caractrisation de lagroindustrie de production damidon aigre de manioc dans le dpartement du Cauca, Colombie. Les cahiers de la recherche dveloppement 1997, 43, 6781. [51] O. Rojas, P. L. Torres, D. Alazard, J. L. Farinet, M. Carmen de Cardoso: Cassava starch extraction: a typical rural agroindustry with a high contamination potential, in Cassava Flour and Starch: Progress in Research and Development. (Ed. D. Dufour, G.M. OBrien, R. Best) CIAT Publication no. 271 Cali, Colombia, 233238, 1996. [52] D. Peters: Agro-processing waste assessment and management in peri-urban Hanoi, Vietnam. J. Sust. Agric. 2004, 25 (1), 6995. [53] T. Nguyen Xuan, T. Nguyen Minh: Research on Technology and equipment for waste treatment in cassava starch processing clusters: Proceeding of a workshop on Cassava post-harvest processing and impact on environment held in Hanoi University of Technology (HUT), Hanoi, Vietnam, May 10-11. 2006, unpublished work. [54] X. Colin, J. L. Farinet, O. Rojas, D. Alazard: Anaerobic treatment of cassava starch extraction wastewater using a horizontal flow filter with bamboo as support. Bioresource Technol. 2006, 98, 16021607. [55] Q. Nguyen Khac: Sweetness from starch: a manual for making maltose from starch, FAO: Roma, Italy. 37, 1996. [56] A. N. Jyothi, K. S. Kiran, B. Wilson, S. N. Moorthy, B. Nambisan: Wet storage of cassava starch: Use of sodium metabisulphite and acetic acid and the effect on starch properties. Starch/Strke 2007, 59 (3-4), 141148. [57] R. C. Marder, A. Graffham, A. Westby: Small-scale starch extraction to improve process efficiency. Natural Resources
www.starch-journal.com