Sie sind auf Seite 1von 16

Optimization of Biodiesel Supply Chains Based on Small Farmers: a case study in Brazil 1.

Introduction

The Brazilian Biodiesel Program is now six years old and has grown considerably, positioning the country as one of the leading producers and consumers of biodiesel fuel in the world. Following a recent legal determination, since 2010 all the diesel fuel sold approximately 45 million cubic meters per year must be mixed with 5% of biodiesel (ANP, 2009). Family farming in Brazil plays a crucial economic role, accounting for around 20% of the national agricultural GDP. In most parts of the country, however, rural workers live and work in poor conditions, with low income and limited access to public resources. In response to this situation, the Brazilian government has singled out social benefits as the main strategic target of the Biodiesel Program, seeking to integrate these small farmers into the production chain as suppliers of raw materials. This strategy can provide better conditions for effective transfer of income and improved living conditions in rural regions. In view of the forecast growth for the coming years, and the lack of adequate facilities, the continued success of the program requires significant investments in structuring the supply chain, to ensure efficient conditions for the production, transportation, and processing of raw materials. It is therefore important to consider the agricultural, industrial and logistic parameters taken as a whole. The purpose of this article is to present an integrated analysis of the supply chain of vegetable oils for the production of biodiesel fuel sourced from family-owned farms, and taking into account the production, transport and crushing of oilseeds, as well as transport of vegetable oils to the biodiesel production units. A mathematical optimization model is proposed, which suggests multiannual investment solutions for the structure of the supply chain as an analysis tool for strategic decision making. The solution includes suggestions for: (i) investments in oilseed crushing units quantity, location, technology, and capacity; (ii) agricultural dimensioning geographic distribution of the production and area occupied; and (iii) logistic planning definition of oilseed and vegetable oil transportation routes. The model design also takes into account the social aspects underlying the Brazilian Biodiesel Program. The optimized solution takes into consideration the social impacts on the region under study, and it is also possible to establish additional constraints, such as the minimum number of producers to be included in the planned chain. Finally, a practical application of the proposed model is presented. The case study was based on an analysis of the conditions required to fulfill the demand for vegetable oil of a Brazilian biodiesel plant. By way of solution, a set of decisions and investments to be made over a ten-year period is proposed, aimed at optimizing the supply chain operation. 2. Biodiesel Fuel in Brazil and Worldwide

The increasing production and use of biodiesel as a replacement for fossil diesel fuel is a global phenomenon, albeit with different motivations. The United States is concerned mostly with establishing an energy mix that is less dependent on imported oil, especially since this commodity typically comes from politically unstable countries. For Europeans, the main motivating factor is to ensure the environmental sustainability of economic growth, through the establishment of goals for replacing fossil fuels and reducing emissions, among others. In the case of Brazil, the strategic focus was on structuring the supply chain around oilseeds produced through family farming, thereby promoting transfer of income and creating employment in rural areas. The social benefits are, therefore, fundamental advantages in the adoption of biodiesel as a fuel, and are defined as one of the pillars that motivate the establishment and the continuity of this program. Studies have shown that for every 1% of fossil diesel that is replaced by biodiesel, 45,000

new jobs are created in rural areas, generating an average annual income of US$ 2,800.00 (HOLANDA, 2004). 2.1. Overview of Global Biodiesel Production The European Union is the world leader in the development of biodiesel production (Figure 1). Germany is the largest producer worldwide, followed by France and Italy. In Europe, biodiesel fuel represents around 80% of the total volume of biofuels produced; that is why it is so important for the establishment of goals and public incentive policies (OCDE, 2006). After the European Union, the United States is the worlds largest producers of biodiesel, with 176 plants currently in operation and production capacity of around 2.61 billion gallons a year. There are a further 39 construction or expansion projects currently being built. Soybean oil is the raw material for over 90% of the volume produced, followed by cottonseed oil, with around 5% (NBB, 2007).

Figure 1: Projections for biodiesel production worldwide. Source: USDA (2008)

It is very difficult to predict the behavior of biodiesel production worldwide over the next ten years. Despite the leadership of the European market, the volume of production planned by the Brazilian government, and in particular, the plans to diversify the supply with castor oil and palm oil, have been attracting international attention. China and India have been making efforts to promote jatropha as a raw material, while other South Asian countries are focusing on biodiesels from palm and coconut oil (OCDE, 2006). 2.2. Overview of Brazilian Biodiesel Production The current offer of oils and fats for the production of biodiesel fuel consists primarily of soybean oil, with 6.26 million tons a year, and cow fat, with 598,000 tons a year (OILWORLD, 2009). In fact, soybean oil currently represents 77.4% of the total volume of biodiesel produced in Brazil, followed by cow fat with 15.5% (ANP, 2009). The installed capacity for biodiesel production in Brazil is significantly higher than the volume effectively produced. There are 63 plants with authorization to operate, as well as 19 awaiting authorization for new installations, and 13 awaiting authorization to extend existing units. Thus, the total authorized production capacity is over 4.5 million cubic meters per annum (ANP, 2009). The current Brazilian legislation states that 5.0% of biodiesel must be added to the mineral diesel sold throughout the country, representing an annual consumption of approximately 2.25 million cubic meters.

The total vegetable oil production capacity in Brazil is around 155,000 tons a day, including both operating and inactive units (ABIOVE, 2009). In Brazil, the industrial park for oilseed crushing is engaged mainly in soybean oil production, reflecting the development of this production chain; the reality for other oilseeds, however, is very different. Most of the oil extraction facilities are located in the South and Central-West regions of the country. In the regions where the federal government has been offering incentives for oilseed cultivation by family-owned farms for the production of biodiesel, the crushing industry is very underdeveloped, especially for units that process the two oilseeds considered a priority: castor bean and sunflower. There are only two crushing units for the production of the first, and none dedicated to the second. To establish an effective plan and suggest production perspectives, we must first comment on certain aspects related to the supply of raw material volumes required to support the scenarios considered. In the case of Brazil, which is seeking to diversify its sources of raw materials, this is even more essential, as the biodiesel industry competes with several other end-consumer segments, such as the food, pharmaceutical, and chemical industries, depending on the vegetable oil in question. The development of the supply of raw materials by family-owned farms for biodiesel production focuses on two oilseeds: castor bean and sunflower. The farms are being continually encouraged to intercrop these with traditional food crops such as beans and corn. This study focuses on the castor oil plant. It therefore gives an analysis of the market for castor bean, oil and cake. 2.3. The Castor Oil Industry in Brazil and Worldwide The leading producer of castor oil worldwide is India, responsible for around 70% of the worlds production. Having been the worlds top producer in the past, Brazil is now the third largest producer, with nearly 8% of the global production. Both countries, together with China, make up 94% of the worlds production (Figure 2).
1600000 1400000 1200000 1000000 800000 600000 400000 200000 0 Brasil Brazil China China India ndia

World Mundo

85

61

65

73

69

77

81

89

93

97

19

19

19

Figure 2: Worldwide production of castor beans, in tons. Source: FAOSTAT (2009)

In general, castor oil is used in all countries of the world, but consumption is greater in the more industrialized countries. France is the top buyer of castor oil, followed by Germany, the United States, Japan and China. Brazil, which from 1978-1982 and 1983-1987 was the top exporter of castor oil overall, reduced its sales considerably in subsequent years; in 2006, it represented only 0.3% of the worlds exports (FAOSTAT, 2009). In 1998, the Brazilian production of castor beans reached the lowest level in twenty years, with only 16,000 tons (IBGE, 2009). Since that year, there has been a noticeable uptrend; 2005 marked the highest production in the past few decades, clearly influenced by the establishment of public policies

19

19

19

19

19

19

19

20

20

01

05

promoting castor oil as a raw material for biodiesel fuel. The 2008 harvest ended with production of 122,000 tons. 3. Family Farming in Brazil

Guaranteeing a regular supply of raw material is an extremely sensitive issue that represents one of the major threats to biodiesel production in any part of the world (IEA, 2002). In some countries, especially those of the European Union, concerns over the expansion of areas occupied by crops for biofuel production, and the possibility these crops compete for space with food crops, have prompted government institutions to regulate the sector by allocating certain percentages of arable lands for the production of oilseeds. In Brazil, the clear strategic option was to establish the involvement of small farmers as one of the supporting bases of the project to produce and use biodiesel. To be included among the suppliers of the Brazilian government program, the small farm must fulfill the following requirements: the area must not be larger than a given size, determined by the government; the labor must be performed predominantly by family members; the income must be obtained predominantly from activities related to the property; the family must live at the property, or within a short distance.

In general, small, family farming is characterized by small, widely dispersed rural properties, and consequently small scales of production. Typically, small farmers have low levels of education and poor living conditions, and use outdated agricultural production methods (GARCIA, 2007). Unlike large, corporate farms, which employ roughly one worker for every 100 hectares of productive land, the small farms have one worker for 10 hectares. In terms of the average size of agriculture and livestock operations in Brazil, around 65% have less than 20 hectares of land, with an average area of 5.7 hectares per enterprise. Small agricultural farms account for nearly 19% of the countrys agricultural GDP, while small livestock production accounts for 13% (DIEESE, 2008). 4. The Brazilian Biodiesel Program and its Challenges

Launched by the Brazilian federal government in December 2004, the National Biodiesel Production and Use Program PNPB is an inter-ministerial program created with the goal of implementing the production and use of biodiesel fuel in a way that is technically and economically sustainable, focusing on social inclusion and on regional development through job and income generation (BRASIL, 2004). Moreover, the federal government created the Social Fuel Seal (SCS), which is granted to industrial biodiesel producers that promote the social inclusion of small farmers through the purchase of minimum quotas of raw materials. With the SCS, these companies can take part in closed biodiesel auctions and receive significant tax exemptions. Contracts with the supplying small farms must include at least one representative of a family farmers union, and the industrial producer must provide assistance and technical training for all small farmers who supply their raw materials (GARCIA, 2007). Seeking to maximize the social benefits, the government selected the so-called semi-arid region characterized by low humidity and low rainfall as a priority for PNPB. The majority of the population in this region lives in rural areas, which have the lowest levels of education, income, and access to healthcare and basic sanitation services in the country. Brazils semi-arid region covers eight of the nine states in the Northeast region, as well as the northern part of the state of Minas Gerais (Figure 3), comprising an area of around 1.0 million square kilometers.

Figure 3: Delimitation of the semi-arid region of Brazil. Source: MIN (2009)

In terms of raw materials for the production of biodiesel, the strategic goal is to diversify, reducing the percentage of soybean in the supply to this industry. For the semi-arid region, the program encourages the production of oilseeds intercropped with grains for the food industry, such as beans and corn, thereby expanding biodiesel production without reducing food production. Due to its hardiness, ability to withstand adverse weather conditions and good adaptation to intercropping, and the widespread knowledge about its production techniques among farmers in the region, castor seed was selected as one of the main oilseeds for biodiesel production in the scope of the PNPB. Despite the predominance of soybean oil as raw material for biodiesel, castor oil has been gaining ground as the principal oilseed acquired by the companies involved in the program. Attempts are also being made with sunflower, but there are some difficulties mainly due to the lack of technical knowledge or tradition in the farming of this crop, particularly in the semi-arid region. The most promising results are being obtained through intercropping, on farms dedicated to fruit crops. In the case of castor oil in particular, although there have been studies on its production and cycle of economic exploration, no clearly established production chain can be identified for this crop as a raw material for biodiesel production. Historically, its volume of production has been subject to wide variations over time, linked to price fluctuations and external demand. This situation has improved since the decision to include castor oil in the biodiesel agro-industrial chain, and progress is forecast in the medium term, through a planned and sustainable structure (AMORIM, 2005). Despite the consensus that the main obstacle to the consolidation of the biodiesel production chain in the semi-arid region consists of guaranteeing a minimum volume, the fact remains that issues relating to the transport logistics of raw materials and biodiesel fuels, and the location of crushing and production units, have not yet been resolved (AZVARADEL, 2008). The installed storage capacity and its conditions represent another weakness in the oilseed logistic chain, particularly in regions with no tradition in this area of production, as is the case of the semi-arid area of the Northeast (VIEIRA, 2006). Biodiesel producers often have to travel long distances to collect small volumes of production. Moreover, almost all of the transportation takes place by road, usually along roads in poor state of repair, which significantly increases the total cost of the product (AZVARADEL, 2008). Studies estimate that bad road conditions increase the overall transport costs of the biodiesel production chain by up to 18% (COPPEAD, 2007). Oilseed from small farmers is purchased at certain meeting points, which are usually located in the areas around rural properties; typically, the farms are located no more than 20 km away from

these facilities. It is, therefore, a logistic operation that is characterized by high capillarity, with hundreds of origins. The agricultural product is always acquired in the form of grain, which must then undergo a processing stage, since the biodiesel plants that adopt the normal transesterification process receive raw materials in the form of vegetable oils. This stage, which consists of extracting the oil from the grain, takes places in dedicated industrial plants known as crushers or extractors. The oilseed crushing stage is a threshold between the farmers and the processing industry, and can represent a critical stage in the financial balance of the production chain. In other words, this echelon adds significant value, and being entrusted with it gives the agent considerable power in relation to the other links of the chain. Vertical integration can translate into economic advantages, as well as guaranteeing the supply of raw materials. It is therefore important to assess the level of interest among biodiesel producers or even producers of vegetable oil derivatives in verticalizing the process in this way. Due to the importance of the oilseed crushing stage in the biodiesel production chain, and the poor conditions of the crushing industrial park in the region targeted by the Brazilian Biodiesel Program, we suggest a mathematical approach as a tool for analyzing and optimizing the structures for supplying vegetable oil to the biodiesel production units. 5. Presentation of the Model and Case Study

The aim of our model is to provide support for strategic decisions related to the definition of the logistic structure, offering a solution for decision making at two levels: the distribution of agricultural production, and the specification of crushing units. The distribution of agricultural production involves selecting specific regions for the production of each type of oilseed, and the occupied area of each region. Basically, the aspects considered are agricultural productivity and production costs, while observing the basic constraint of area available for cultivating the crop in each region. The distribution is determined on a yearly basis, throughout the period of analysis. It is important to emphasize that the determination of distribution of the agricultural production is sensitive to social aspects in that it takes into account the number of families involved in each of the regions selected. It is possible to apply additional constraints related to social inclusion strategies, such as establishing a minimum number of families included for each year of the analysis. For the specification of crushing units, four parameters are determined: the number of facilities, the technology selected for each facility, the geographic location, and the scale of the processing capacity. The best time to make investments in each unit should also be determined, considering each year throughout the period of analysis. The final aspect in determining the structure is to indicate transport routes for the oilseeds and vegetable oils. For each year of the period of analysis, the origin and destination points are defined for all the movements of material necessary to supply the biodiesel production unit with vegetable oil with oilseed acquired from small farmers. The model elaborated in this study was applied to a small group of production zones, and considers only castor oil production. The case study focuses on the supply of vegetable oil to a single biodiesel plant. Nonetheless, the structure of the model was designed in such a way that it can be applied to any other region, regardless of its geographic size, or to other oilseeds and production plants; the mathematical formulation does not change, and the scale of the model can be expanded as required. Generally, the problem of determining the best location for production zones and crushing units is to find an optimum point between the two extremes, seeking to balance logistic costs with investment costs. At one extreme, only one very large crushing unit would be built (minimum investment costs); at the other, a large number of very small units would be built (minimum logistic costs). Therefore, the model can be described in schematic form as a tool for minimizing a mathematical function of the total supply cost, subjecting the solution to a set of constraints and

considering a set of parameters. The output consists of sets of values for the variables that determine the optimum solution, as shown in Figure 4.
Input data
Average property Crop yield Production cost Area available Oil content Oilseed transportation cost Oil transportation cost Capacity Efficiency Crushing cost Installation cost

Objective function
Agricultural costs Industrial costs Logistic costs

Constraints
Production demand Social demand Crushing capacity

Demand Number of families

Decision variables
Productive area Families Installation decision Amount of crushed oilseed

Amount of transported oilseed Amount of transported oil

Volume purchased in the market

Figure 4: Representation of the model for the case study, in schematic form.

The input data can be grouped in sets of four different types: agricultural (related to oilseed production), industrial (related to the industrial crushing facilities), logistic (related to transportation), and demand-related (involving the biodiesel plant and social aspects). The mathematical formulation itself consists of an objective function to be minimized, which considers the costs involved in the operation of the chain (agricultural, industrial and logistic), as well as sets of constraints associated with the demands to be fulfilled. Finally, the solution for the set of linear equations described above, with the input data considered, provides values for the decision variables which, like the input data, can also be classified for didactic purposes into four distinct types. Initially, the sets of indices of the model, to which all the other data would refer, were determined, namely: production zones, locations of the crushers, types of crushers, oilseeds, and production plants. Production zone (p): territories determined according to the discretization of the universe to which the study is restricted; they can be as small as the desired level of granularity for the model. In the present study, the micro-regions located in the northern area of the state of Minas Gerais, comprising around 220,000 square kilometers, were defined as production zones. Two criteria were observed: proximity to the district where the biodiesel plant is located, and the geographic boundaries of the Brazilian semi-arid region. Oilseed (o): crops cultivated for the purpose of extracting vegetable oil for biodiesel production. Based on its agricultural potential for the region of interest, castor oil was selected as the only oilseed for the development of this work. The strategic alignment between incentives offered by biodiesel producers and by the Brazilian government was also taken into account. Crusher type (k): the possible technological configurations for the crushing unit to be installed. Of the existing oilseed crushing technologies, mechanical press and solvent extraction were selected. Also, different capacities were determined for each of these technologies, in order to represent small, medium and large units. Crushing plant possible location (c): the districts with the minimum required conditions for the sustainable installation of an oilseed crushing unit. Districts are selected which have good infrastructure, relatively close to the production zones, and which are strategically located close to paved highways in good conditions.

In the region under study, the districts considered suitable for the installation of a crushing unit were those with a population of over 40,000 inhabitants. Biodiesel Plant (u): units that produce biodiesel from vegetable oils or cow fat, through the process of transesterification. The model presented here evaluates the elaboration of an arrangement for supplying a single biodiesel plant, located in the district of Montes Claros in the state of Minas Gerais. Time period (t): the analysis horizon of the model, which is 10 years. Scenarios (s): as this problem relates to agricultural productivity, which always involves a high level of uncertainty, the formulation of a stochastic model was selected. Therefore, the productivities are specified in relation to three scenarios: pessimistic, realistic, and optimistic. The pessimistic scenario represents an unexpected interruption to the harvest, caused by external factors such as prolonged droughts, causing a drop in production. The optimistic scenario reflects a situation in which the harvest is influenced by some factor, usually climatic, which leads to an increase in productivity of a given oilseed. Based on these initial definitions, the data to be collected were determined. These can be organized according to the echelon of the production chain with which they are associated: agricultural production, crushing, or transportation of oilseeds and oils (the most relevant data can be seen in the appended tables). There are also some additional constraints that can be controlled by the decisionmaker, using the model as a tool for optimization. The objective function of the model represents the total costs of the supply chain, and the optimized solution consists of its minimization given the various constraints imposed. In short, the function is comprised of the following factors: oilseed and oil transportation costs, agricultural production costs, oilseed processing costs at the crushing units, investment costs related to the installation of the unit, and the costs associated with the purchase of any additional volumes of oil in the market, to meet the demand of the biodiesel plants. The formulation can be seen below.

As for the problem of location, the optimum solution to the system of equations provides sets of values for decision-making variables. The main ones are: areas used annually for crops in the micro-

regions with areas available; number of families involved annually in each of these micro-regions; specification (technology and size) and location of the crushing units throughout the analysis period; profile of the supply of raw materials to the crushers (oilseed flows) and to the plants (oil flows); and volumes of oil acquired in the market, in addition to the inherent total cost of the system. Table 1 lists the parameters that make up the formulation, with their brief descriptions. Table 2 lists the decision variables to be determined.
Table 1: Characterization of the parameters of the model Parameter Description percentage of oil in the oilseed O Efficiency of crusher type K. Ratio between the amount of oil obtained through crushing and the total oil mass contained in the seeds Crushing costs for crusher type K. Total expenses involved in the industrial crushing process, excluding costs with raw materials. Includes input, labor, maintenance, and equipment depreciation Annual capacity of crusher type K. Maximum amount of oilseeds that can be processed per year by an industrial crushing unit. The following capacities were considered: 12,500, 25,000, 62,500, 125,000, and 200,000 Probability that scenario S occurs Total available area production zone P at period T oilseed O productivity at production zone P, under scenario S, and at period T. Volume of grains harvested in the cultivated area. Historical data on the productivity of castor oil crops (combined with corn or beans) were considered in the regions under study oilseed O production cost at production zone P. Total expenses related to the production, including labor, inputs and technical assistance. Data published by departments of Brazils Ministry of Agriculture were considered average land size at production zone P at period T User-controlled parameter that allows the user to define an additional constraint related to a minimum number of families to be hired for oilseed supply during period T Installation costs for crusher type K at period T. Total investments required to install a crushing unit. Varies annually along analysis horizon Transportation cost between I and J; namely, production zone P and crushing plant possible location C vegetable oil cost for biodiesel plant U at period T Demand by biodiesel plant U at period T. Annual demand for vegetable oil supply by the biodiesel plant

Table 2: Characterization of the variables of the model Variable Description Decision to install a K-type crushing plant at location C, and at period T Area selected for the production of oilseed O at production zone P, scenario S, period T Number of farming families producing at productiona zone P, scenario S, period T Amount of oilseed O transported from I ( production zone P) to J (crushing plant possible location C), scenario S, period T Amount of oilseed O crushed at K-type crushing plant, location C, scenario S, period T Volume purchased in the market to fulfill demand of biodiesel plant U, scenario S, period T Auxiliary variable for cumulative investments

6.

Results of the Case Study

As mentioned earlier, the solution to the problem addressed consists of a decision at two levels, which involves defining the location and type of crushing units and the distribution of the agricultural production. The analysis horizon is ten years, and the solution is multi-annual in relation to all the decisions. The base scenario considered is that which is determined by the set of constraints and parameters, considering the variability of the agricultural productivity. For these conditions, optimized distribution of production is widely dispersed in the first two years, but over the long term a tendency toward centralization around production centers is observed, particularly for the three production zones of Araua, Capelinha and Pedra Azul (Figure 5b). The dispersion of production in the first years is due to the lack of available area, which is insufficient to meet all the demand from the plant. Fourteen out of the sixteen zones were involved in oilseed production in the two first years (Figure 5a). In the third year, five of them had stopped producing, and from the fifth year onwards, only four Araua, Capelinha, Pedra Azul and Janaba were still part of the oilseed supply chain (Figure 5b).

Figures 5a and 5b: Areas earmarked for the cultivation of castor oil by productive zones in 2009 and 2018.

The first level of decision-making is to recommend the type and location of the crushing units that will process the oilseeds produced and subsequently supply the biodiesel plant. For this scenario, the optimized solution recommends the installation of three crushers; the first with capacity for 25,000 tons of oilseed, built in 2009, the first year of the analysis, in the district of Montes Claros. In the

following year, 2010, Araua would be the site of installation of a unit with capacity for 12,500 tons of oilseed. Finally, in 2017, a unit similar to the one in Araua would be installed in the district of Capelinha. The technology adopted in all cases is the mechanical press, which has lower operating and investment costs but is also less efficient than solvent extraction units. This choice is justified by the fact that the units selected are small and medium in size, and do not provide enough economy of scale to offset the higher fixed and investment costs required for solvent extraction. After validating the model, new simulations are performed, considering different sets of parameters that represent possible scenarios for the activities studied. The purpose of these sensitivity studies is to increase the robustness of the solutions obtained, from the moment they become recurrent despite certain variations in the values of the parameters. Some alternative scenarios for study are shown in Table 3.
Table 3: Scenarios for sensitivity analysis. Scenario Variation (-50% and +50%) in freight costs Variation (+50% and +100%) in demand Variation (-15% and +30%) in investment costs Justification The structure of the oilseed-collecting activities is still under development, and the costs still cannot be precisely associated to market values. These variations represent the possibilities for expanding the biodiesel plant capacity. Typical uncertainties involved in budget estimations during preliminary stages, as in this case.

Table 4 summarizes the solutions suggested for the types of crushers and installation locations over the period of analysis, for the base scenario and its variations referring, in each case, to mechanical press technology. Capacities are expressed in tons of oilseed per year.
Table 4: Location and installed capacity of the crushers. Scenario Base case Investment cost 15% lower Investment cost 30% higher Location of the crusher plants Montes Claros Capelinha Janaba Montes Claros Demand 50% higher Araua Capelinha Montes Claros Demand 100% higher Paracatu Araua Capelinha Freight cost 50% lower Montes Claros Montes Claros Freight cost 50% higher Araua Una Capelinha 62.500 12.500 12.500 12.500 12.500 62.500 12.500 25.000 12.500 62.500 25.000 12.500 2009 25.000 12.500 25.000 12.500 2010 2012 2017

In general, the crushing units would be located in regions close to where the production is distributed, thereby creating self-sufficient oilseed production and oil crushing centers. The only exception is Montes Claros, which does not have long-term occupied areas due to their low productivity, and its crusher would be supplied with oilseeds from other production zones. Regarding the technology, it is seen that mechanical press units are the best option in every case. Solvent extraction crushers achieve better extraction efficiency, but require higher initial

investments. Typically, mechanical press units are preferred for facilities with smaller capacity, and solvent extraction units are preferred for larger facilities that can achieve significant economy of scale. Analysis of the sensitivity of the solution to freight prices clearly demonstrates the trade-off between investment costs and transport costs. The cost of transporting vegetable oil is considerably lower than that for grains, mainly due to the difference in volumes transported. With rising freight prices, the tendency is to install more crushing units in order to reduce the total volume of grain to be transported. Similarly, when freight prices are lower, there is the opposite tendency to centralize the crushing process, leading to a larger volume of transported grain. The objective function the equation minimized in the solution of the model has a total value of approximately US$ 200 million, of which around US$ 6.3 million are invested in industrial crushing facilities. This amount includes costs related to agricultural production, transportation, installation and operation of the crushing units, giving an estimated value of around US$ 1,630.00 per ton of castor oil produced and delivered to the biodiesel plant. 7. Conclusions

Brazil has been attracting worldwide attention as a major producer of biodiesel, achieving considerable growth in recent years. The Brazilian government has established social benefits as one of the pillars of its biodiesel program, in an attempt to reduce poverty and transfer income to rural areas by including small farmers in the oilseed supplier base for biofuel production. The continued success of this program requires considerable investments in production and logistic structures. The key link in this supply chain is the oilseed crushing units, particularly due to the poor current conditions of these facilities in the country. Consolidated agricultural production chains, or those related to large farming operations, have been the object of a considerable number of publications. Less structured production chains, such as that of castor oil, however, are at most the object of qualitative analyses, or studies restricted to sociological or macro-economic aspects. The use of mathematical tools to support decision-making at a strategic level plays an important role in ensuring the economic feasibility of the investments and the sustainability of the business as a whole. In this regard, this article proposes a mathematical model to represent this production chain and help determine an optimized structure for the supply of vegetable oils to biodiesel plants, considering aspects related to the costs of agricultural production, transport of oilseeds and oils, and industrial investments in the installation of crushers and their processing costs. A practical application of this model is described for oilseed supply to an existing biodiesel plant in Brazil. In this case study, there is a noticeable tendency to centralize the agricultural production and crushing process, preferably combining both stages in the same region. There is also a tendency to centralize the crushing process in the region close to the biodiesel plant. In each case, the solution provided by the model is to invest in mechanical press units rather than solvent extraction technology. This is due to the high initial investment of the latter, without the economy of scale gained through the higher volume of oil obtained by increased efficiency. Given that the mathematical model was designed to be applicable to various scenarios, its application to a wider analysis is suggested as a topic for further study, considering more than one type of oilseed and more than one biodiesel plant, and with the option of selling off surplus production of vegetable oil. The formulation of the model as it is presented can be a useful decision-making tool for a biodiesel company seeking to establish a supply chain based on family agriculture. Even in the case of a simple verticalization process that takes in the crushing echelon, it is interesting that this decision is made through an iterative process, taking into account the geographic definition for seed distribution. Considering agricultural parameters such as available area, production costs and agricultural productivity, industrial parameters such as installation costs, technical indices and processing costs, and

logistic parameters such as oilseed and oil transport costs, all of which interrelated, gives a solution with much higher added value than by taking decisions on these aspects in isolation. Finally, we encourage other academics to work on numerical and quantitative approaches to production chains based on the activities of small farmers. Bearing in mind especially the recent nature of the Brazilian Biodiesel Program, any idea that that economic feasibility and family farming are unrelated in terms of agro-energy must be firmly rejected. Thus, academic work on this subject can play a fundamental role in helping to ensure that the strategic goals for social inclusion and income transfer to rural areas are achieved. 8. Bibliography

AMORIM, P. Q. R. Perspectiva Histrica da Cadeia da Mamona e a Introduo da Produo de Biodiesel no Semiarido Brasileiro sob o Enfoque da Teoria de Custos de Transao. Monografia. USP/Esalq. Piracicaba-SP, 2005. ABIOVE (Associao Brasileira das Indstrias de leos Vegetais). Disponvel em http://www.abiove.org.br. Acesso em 20/12/2009. ANP (Agncia Nacional do Petrleo). Boletim Mensal de Biodiesel. Disponvel em http://www.anp.gov.br/biocombustiveis/biodiesel.asp. 2009. AZVARADEL, A. C. A Contribuio da Poltica Estadual para Viabilizar a Participao da Agricultura Familiar no Programa Nacional de Produo e Uso do Biodiesel: o Caso da Bahia. Dissertao de mestrado. UFRJ. Rio de Janeiro, 2008. BRASIL. Programa Nacional de Produo e Uso do Biodiesel. Braslia, 2004. COOPEAD. Planejamento Estratgico Tecnolgico e Logstico para o Programa Nacional de Biodiesel. Relatrio tcnico. Disponvel em <http://www.ibp.org.br>. 2007. DIEESE. Estatsticas do Meio Rural 2008. Braslia, 2008. FAOSTAT(Food and Agriculture Organization of the United Nations). Disponvel em <http://faostat.fao.org/site>. Acesso em 20/12/2009. FRIEDRICH, S. A World Wide Review of the Commercial Production of Biodiesel A technological, economic and ecological investigation based on case studies. Relatrio tcnico. Wirtschafts Universitt. Viena, 2004. GARCIA, J. R. O Programa Nacional de Produo de Biodiesel Brasileiro e a Agricultura Familiar na Regio Nordeste. Dissertao de mestrado. UNICAMP. Campinas, 2007. HOLANDA, A. Biodiesel e Incluso Social. Braslia, 2004. IBGE (Instituto Brasileiro de Geografia e Estatstica). Disponvel em <http://www.sidra.ibge.gov.br>. Acesso em 20/12/2009. IEA (International Energy Agency). The Development of Biodiesel in Germany. Relatrio tcnico. Bioenergy Task 27 Liquid Biofuels. Viena, 2002. MIN (Ministrio na Integrao Nacional). Mapa do semirido. Disponvel em <http://www.integracao.gov.br>. MONTEIRO, J. M. G. Plantio de Oleaginosas por Agricultores Familiares do Semiarido Nordestino para Produo de Biodiesel como uma Estratgia de Mitigao e Adaptao s Mudanas Climticas. Tese de doutorado. UFRJ. Rio de Janeiro, 2007. NBB (National Biodiesel Board). Benefits of Biodiesel. Disponvel em <http://www.biodiesel.org/resources/fuelfactsheets>. 2007. OILWORLD. Statistics. Disponvel em <http://oilworld.biz>. Acesso em 27/04/2009. UFOP. Biodiesel Production and Marketing in Germany: The Situation and Perspective. Relatrio tcnico. 2002. USDA (United States Department of Agriculture). Global Agricultural Supply and Demand: Factors Contributing to the Recent Increase in Food Commodity Price. Disponvel em http://ers.usda.gov. 2008.

VIEIRA, J. N. S. A Agroenergia e os Novos Desafios para a Poltica Agrcola no Brasil. O Futuro da Indstria: Biodiesel. Braslia, 2006. 9. Appendices

This section presents some tables with real data considered in the elaboration of the case study described in this article.
Table 5: Selected production zones Production zone Almenara Araua Capelinha Bocaiva Gro Mogol Janaba Januria Montes Claros Total area (ha) 1,545,200 1,026,100 1,201,200 789,600 907,600 1,515,500 3,316,900 2,224,800 Population 175,991 154,850 196,894 66,040 42,046 243,080 269,999 592,000 Production zone Pirapora Salinas Nanuque Tefilo Otoni Diamantina Pedra Azul Paracatu Una Total area (ha) 2,307,200 1,783,700 847,200 1,160,900 734,800 506,900 3,499,700 2,738,400 Population 159,963 208,739 118,587 258,080 82,840 84,313 208,367 142,122

Table 6: Selected crushing plants possible locations Brasilndia de Minas Paracatu Bocaiva Janaba So Francisco Corao de Jesus Mirabela Buritizeiro Rio Pardo de Minas Nanuque Almenara Capelinha Joo Pinheiro Una Gro Mogol Januria Braslia de Minas Francisco S Montes Claros Pirapora Diamantina Tefilo Otoni Araua

Table 7: Considered yields for each production zone (base case, first year of analysis) Production zone Una Paracatu Januria Janaba Salinas Pirapora Montes Claros Gro Mogol Yield (ton/ha) 1,462* 1,462* 1,149 1,395 1,065 1,229 799 1,265 Production zone Bocaiva Diamantina Capelinha Araua Pedra Azul Almenara Tefilo Otoni Nanuque Yield (ton/ha) 968 800 2,000 1,462* 1,519 1,462* 1,462* 1,462*

* Average yield at the state of Minas Gerais

Table 8: Total production cost for each production zone Production zone Una Paracatu Januria Janaba Salinas Pirapora Montes Claros Gro Mogol Total production cost (US$/hectare) 663.95 664.07 665.83 669.26 670.25 665.67 666.13 666.12 Production zone Bocaiva Diamantina Capelinha Araua Pedra Azul Almenara Tefilo Otoni Nanuque Total production cost (US$/hectare) 668.02 668.31 669.94 665.51 663.31 664.06 666.00 665.05

Table 9: Available areas at each production zone (ha) Production zone Una Paracatu Januria Janaba Salinas Pirapora Montes Claros Gro Mogol Bocaiva Diamantina Capelinha Araua Pedra Azul Almenara Tefilo Otoni Nanuque Available area year 1 1,221 988 2,453 909 1,655 1,898 1,185 406 310 333 566 992 219 383 326 124 13,968 year 2 2,148 1,739 4,317 1,599 2,912 3,340 2,085 714 545 586 996 1,745 386 674 574 219 24,579 year 3 3,781 3,061 7,599 2,814 5,125 5,879 3,670 1,256 959 1,032 1,754 3,072 679 1,186 1,010 385 43,262 year 4 6,654 5,387 13,374 4,953 9,020 10,347 6,460 2,211 1,688 1,816 3,086 5,407 1,194 2,087 1,778 678 76,140 year 5 11,711 9,481 23,538 8,717 15,875 18,211 11,369 3,892 2,972 3,196 5,432 9,515 2,102 3,674 3,129 1,193 134,007 years 6 to 10 12,205 9,881 24,531 9,085 16,545 18,979 11,849 4,056 3,097 3,331 5,661 9,917 2,191 3,829 3,261 1,243 139,661

Table 10: Oil content Oilseed castor seed Oil content 45%

Table 11: Industrial efficiency for each crushing plant type Efficiency in % Mechanical press 90.4 Solvent extraction 96.3

Table 12: Total crushing cost for castor oil production (US$/ton) Capacity (t seed / year) 12,500 Mechanical press Solvent extraction 142.86 200.57 25,000 108.57 157.71 62,500 81.71 123.43 125,000 69.71 108.57 200,000 65.14 103.43

Technology

Table 13: Crushing plant investments (million US$, first year of analysis) Capacity (t seed / year) Technology Mechanical press Solvent extraction 12.500 4.057 5.257 25.000 6.343 8.229 62.500 9.314 12.114 125.000 14.571 19.143 200.000 19.829 25.600

Table 14: Freight cost (US$/t.km) Material oilseeds oil Cost (US$/t.km) 0.1253 0.1810

Table 15: Biodiesel plant demand (t oil / year) Biodiesel plant Montes Claros Demand (t oil / year) 15,000

Das könnte Ihnen auch gefallen