Tropentag 2007 University of Kassel-Witzenhausen and University of Göttingen, October 9-11, 2007 Conference on International Agricultural Research for Development Economic analysis of different processing methods for small-scale coconut oil production in the Philippines Matthias Kindermann a*, Kevin Weis a, Christian Lippert b, Joachim Müller a a Universität Hohenheim, Institute of Agricultural Engineering, Garbenstraße 9, 70599 Stuttgart, Germany. b Universität Hohenheim, Institute for Farm Management, Schwerzstraße (Torbogen 3), 70599 Stuttgart, Germany. Introduction In the Philippines the population has grown by almost 13 % from the year 2000 to 2005. As the majority of the people still cook with traditional fuels like wood or charcoal the demand for these combustibles is rising, too (www.census.gov.ph, 2006). Apart from serious ecological problems like deforestation which are linked to this development, increasing fuel prices are a big problem especially for the poor people. A plant oil stove which was developed by the Bosch and Siemens Home Appliances Group could be a possible alternative. Crude coconut oil which can be produced locally on small scale level is supposed to serve as fuel for the stove. To produce coconut oil on household level the kitchen methods (also called wet methods) are applicable possibilities. The raw material is the fresh coconut. A second possibility is to establish the dry method on cooperative level. Here the oil is extracted with the help of an expeller press from the dried coconut meat that is called copra. The copra is bought from farmers/dealers or it is obtained from fresh nuts by the cooperative itself. In this paper it was analysed whether the small scale coconut oil production can be profitable and if coconut oil can be locally produced to a price which enables to compete with the common cooking fuels like kerosene, LPG, fuel wood and charcoal. Material and Methods Wet methods Using the wet method the grated coconut is mixed with water and then squeezed with a bridge press or using the hands to obtain a coconut oil-water-emulsion. Three different ways to extract the oil out of the emulsion were analysed: Traditional kitchen method (TKM): The emulsion is cooked to separate the oil. Modified kitchen method (MKM): The emulsion is kept in a container for at least two hours, where a settling process takes place. The water and the residues are placed in another container and only the milk is cooked to extract the oil. Virgin Oil Method (VOM): The emulsion is kept in a container for settling and after two hours the separated water is removed. This has to be repeated after four hours. During the storage in the container a fermentation process takes place. After 18 48 hours three phases are separated: 1. coconut cream at the top (it still contains oil which can be * Corresponding author. Email: makinder@uni-hohenheim.de
extracted through heating and be used as cooking oil) 2.Virgin Oil in the middle and 3. waste residues at the bottom (SABUSAP, 2005) Dry method The performance of three expeller presses (German made Strähle SK 60/2 and Montforts Komet DD 85 G and Philippine made Simplextractor) that can be used on cooperative level was analysed. It was distinguished between Case I processing copra and Case II processing fresh nuts. For the economic analysis the following key figures were calculated: Investment costs for coconut oil production centre, production unit costs for one litre of oil (REISCH and KNECHT 1995), payback period (REISCH and ZEDDIES, 1992) and net present value (NPV) (BRANDES and ODENING, 1992). Further on a sensitivity analysis (BREALEY et al, 2006) and an analysis of the coconut price development were done. In all scenarios linear depreciation was utilised. If there was no data available regarding repair and maintenance costs a certain percentage (3-10 %) of the investment was assumed to represent these amounts. Results Raw data regarding the technical performance of the different oil processing methods was provided by the Leyte State University. As a benchmark for the production costs the kerosene price (36.08 PHP/l in Baybay, Leyte 11/2005) was chosen because the functional principle and the used fuel of the kerosene and the plant oil stove are both similar. Whenever revenues from the oil production had to be calculated a producer price of 30 PHP/l was assumed. Investment Price Product life Maintenance costs Maintenance costs (% on investment) (PHP) (a) (PHP/a) Simplextractor 214,596 5 10,730 5 Montforts 648,670 20 32,434 5 Strähle 658,905 20 32,945 5 Hammer mill 50,000 15 2,500 5 Settling, filtering, storage unit 9,500 15 475 5 Bridge press 1) 36,000 15 3,600 10 Coconut grater 1) 10,000 10 1,000 10 Plant oil stove 1) 2,500 10 250 10 Pot 1) 650 10 65 10 Small filtering unit 1) 1,000 10 100 10 Machete 300 10 30 10 Dehusking tool 275 10 28 10 Scooping tool 250 10 25 10 Coconut convection dryer 150,000 15 7,500 5 Processing centre 444,300 20 44,430 10 Small processing shed 1) 60,000 15 6,000 10 1): for wet methods only Table 1: Investments to start oil production, product lives and maintenance costs
The coconut price in Baybay, Leyte was 2.75 PHP/nut in October 2005. As the production centre will be located in the producing area transport costs to the city can be saved. Thus, all calculations are based on a coconut price of 2.60 PHP/nut. All other mentioned prices refer to the market situation in Leyte in October 2005. An enumeration of all investments which are necessary for wet and dry methods is shown in Table 1. Wet methods The assumption was made that oil is produced out of 30 nuts on 260 days per year. An initial investment of about 110.000 PHP (Table 2) will be necessary. The TKM is very labour intensive so that 300 min are needed to achieve a net oil yield (assuming a plant oil stove is used when cooking is required) of only 0.18 l per 30 nuts. The filtered oil production of the MKM and the VOM are similar with 1.95 l, respectively 1.91 l per 30 nuts. But due to the fermentation process and as no cooking is needed the VOM gives a higher net oil yield with 1.91 l compared to 1.32 l for the MKM. VOM required the lowest working time with 130 min per 30 nuts. Using the MKM 180 min are needed to process the same amount of nuts. The NPVs are negative for all three possibilities ranging from -195,000 PHP (VOM) to -471,000 PHP (TKM). Method Investment* NPV** Payback period (a) TKM 112-471 - MKM 112-279 - VOM 109-195 - * 1 = 65 PHP (10/2005) ** project life 10 a; interest rate 7 % Table 2: Economic key figures of wet coconut oil production methods The production unit costs (including revenues from by-products) are 1,253 PHP/l for the TKM, 136 PHP/l for the MKM and 83 PHP/l for the VOM. The NPVs for all methods are negative with the assumed prices and setup. The sale of the by-products coconut husk, coconut shell and squeezed coconut meat is important to reduce the production unit costs of the oil. The structure of the unit costs is shown in Table 3. Method TKM MKM VOM (PHP/l) (PHP/l) (PHP/l) Coconuts 433.33 58.65 40.84 Labour 614.58 49.91 24.33 Electricity 9.93 1.34 0.94 Repair & maintenance 238.73 32.31 21.86 Machinery 157.36 21.30 13.93 Building 182.54 24.70 17.20 Unit Cost 1,636.47 188.21 119.09 - Husk 25.00 3.38 2.36 - Shell 50.00 6.77 4.71 - Squeezed coconut meat 308.89 41.80 29.11 Unit Costs including byproducts revenues 1,252.58 136.25 82.91 Table 3: Costs per litre coconut oil using wet methods
In all three cases raw material and labour cause the highest costs per produced litre of oil. With regard to the high unit costs and the negative NPVs it isn t economically feasible and therefore not recommendable to produce oil as a fuel on household level with any wet method. If nevertheless due to unavailability of any other fuel one of these methods should be implemented it is advisable to choose the VOM. Dry methods For oil production with expeller presses 260 production days with 8 working hours were assumed. The following technical characteristics of the expeller presses were determined: Strähle SK 60/2 is constructed to extract oil from all kind of oil seeds attaining a capacity of 15 kg/h and an average settled oil recovery of 49.9 % using copra. Montforts Komet DD 85 G is also constructed for all kind of oil seeds showing for copra a capacity of 35 kg/h and an average settled oil recovery of 46.9 %. Simplextractor is mainly used for processing copra. It has a capacity of 39 kg/h and an average settled oil recovery of 51.8 %. It was distinguished between Case I processing copra and Case II processing fresh nuts. The installation of a coconut dryer proved to be necessary as bought copra is often of poor quality and has to be dried further in order to reach a moisture content of 3-7 % which is necessary for an efficient expeller performance. Especially during rainy weather periods the common sun drying is not possible. When processing fresh nuts to copra with the help of a natural convection dryer, high quality copra with few aflatoxin and Poly-Cyclic-Aromatic-Hydrocarbons content can be produced. The investment cost for a coconut oil processing centre ranged from 1,381,000 PHP implementing a centre using the Strähle Expeller to 937,000 PHP with a setup including the Simpextractor (Montforts: 1,369,000 PHP) (Table 4). Expeller Investment NPV* Payback period** (a) Case I Case II Case I Case II Strähle 1,381-2,916-1,615 - - Montforts 1,369-788 560-6 Simplextractor 937 630 2,411 4 2 * 1 = 65 PHP (10/2005) ** project life 10 a; interest rate 7 % Table 4: Economic key figures of small coconut oil production methods The production unit costs (including revenues from by-products) in Case II amount to 44.11 PHP/l, respectively 48.63 PHP/l in Case I using the Strähle. 28.71 PHP/l in Case II, respectively 33.50 PHP/l in Case I can be attained with the Montforts and 23.58 PHP/l in Case I, respectively 28.65 PHP/l in Case I using the Simplextractor (Table 5). The sales of the byproducts proved to reduce the unit costs decisively: Processing copra the costs are reduced on an average of ~10 % and processing fresh nuts on an average of ~17 %. NPVs for the Strähle are all negative. The Montforts has a negative one in Case I and a positive one in Case II with 560,239 PHP resulting in a pay back period of six years. The best and safest investment is the Philippine machine with a NPV of 629,566 PHP in Case I (pay back period four years) and a NPV of 2,411,437 in Case II (pay back period two years) (Table 4).
Expeller Strähle Montforts Simplextractor (PHP/l) (PHP/l) (PHP/l) (PHP/l) (PHP/l) (PHP/l) Case I Case II Case I Case II Case I Case II Passa 1 22.20 23.66 21.88 Coconuts 18.25 19.45 17.36 Electricity 2.11 2.11 1.45 1.45 0.98 0.98 Salary fix 10.29 10.29 4.70 4.70 3.76 3.76 Salary nut 2.11 2.24 2.00 preparation Coconut husks 0.14 0.14 0.15 Repair & 5.44 5.44 2.46 2.46 1.56 1.56 maintenance Machinery 8.33 8.33 3.76 3.76 2.10 2.10 Building 3.95 3.95 1.81 1.81 1.45 1.45 Unit cost 50.48 50.48 37.97 35.87 31.88 29.21 - Copra cake 3.84 3.84 4.47 4.47 3.23 3.23 - Husk 0.42 0.45 0.40 - Shell 2.11 2.23 2.00 Unit costs 48.63 44.11 28.71 28.65 23.58 including byproducts revenues 1: copra at farm gate level with moisture content of ~20 %; 11.50 PHP/kg (Baybay, Leyte 10/2005) Table 5: Cost distribution per litre coconut oil using dry methods Coconut prices and sensitivity analysis Figure 1 illustrates the high fluctuation of the coconut prices in Leyte. In 1995 the average farm gate price was 1.40 PHP/nut. Within the next four years the average annual price increased to 3.24 PHP/nut in 1999. Only two years later in 2001 the mean price fell to only 0.80 PHP/nut. The highest price was attained in 2004 at an average of 4.14 PHP/nut. The mean price from 1999 until 2004 is 2.11 PHP/nut. 4 Coconut Price (PHP/Nut) 3 2 1 0 1996 1998 2000 2002 2004 Year Source: derived from Philippine Coconut Authority data, 2005 Figure 1: Coconut farm gate prices in Leyte 1994-2004
A sensitivity analysis was done for three coconut price scenarios: best (1.60 PHP/nut), actual (2.60 PHP/nut) and worst (4.60 PHP/nut). All three wet methods had negative NPVs, even in the best case situation. Regarding the dry method the Strähle attained only negative NPVs ranging from -714,000 PHP for the best case to -3,417,000 PHP for the worst case. The Monforts showed a profitable performance for the actual case and the best case with NPVs of 560,000 PHP, respectively 2,663,000 PHP. Regarding the Simplextractor the NPVs were also positive for the actual case and the best case (2,411,000 PHP, respectively 4,755,000 PHP). But even this obviously most profitable setup had a negative NPV for the worst case with -2,275,000 PHP (Table 6). This result underlined the crucial influence of the raw material prices on the competitiveness of the oil production. A sensitivity analysis assuming the purchase of copra as raw material showed similar results. Simplextractor is still the safest option followed by the Montforts expeller. Expeller Case Coconut price Unit costs when accounting for by-products resulting NPV Pay back period (PhP/nut) (PhP/l) (1,000 PhP) (a) Strähle best 1.60 37.09-714 - actual 2.60 44.11-1,615 - worst 4.60 58.15-3,417 - Montforts best 1.60 21.23 2,663 2 actual 2.60 28.71 560 6 worst 4.60 43.67-3,645 - Simplextractor best 1.60 16.90 4,755 1 actual 2.60 23.58 2,411 2 worst 4.60 36.93-2,275 - Table 6: Influence of the coconut price on unit costs, NPV and pay back period when using dry methods Conclusion and Outlook In all cases the raw material price has a big influence on the unit costs because it accounts for the biggest part of the expenditures. If the assumed expeller capacities cannot be fully used this will also affect the profitability of the oil production negatively. Modifications of the setups might improve the profitability of all regarded processing methods. With the assumed setup for the dry method it is economically feasible to install a coconut oil processing centre with a Simplextractor press on cooperative level. This solution causes the lowest unit costs, attains the highest NPV and has the shortest period of amortisation. Also the availability of spare parts would be rather guaranteed than for the German machines. Even including a margin for the producer, transportation and distribution costs the produced oil should be competitive with e.g. the kerosene price which was 36.08 PHP/l at the CALTEX gas station Baybay, Philippines in November 2005. Processing fresh nuts is the favourable solution as the revenues from the by-products make the project more profitable. With a rising coconut oil price or decreasing raw material prices the Montforts Komet DD 85 G might be an alternative. The coconut oil production on household level using the wet methods is not economically feasible if the produced oil is used as a fuel for a plant oil stove. Only the VOM can be profitable in case the oil is sold as virgin oil on the local market. The combination of the wet methods producing virgin oil on household level and the dry method on cooperative level extracting oil from the squeezed coconut meat which will then be used as fuel for a plant oil stove might be the most profitable solution for all people involved. Nevertheless it would still require an economic analysis to prove this hypothesis.
References www.census.gov.ph, 2006 (actualised 31.January 2006) Philippines in figures. National Statistic Office, Republic of the Philippine. SABUSAP, ANGEL T., College of Engineering and Agri-Industries, Leyte State University, August 2005 REISCH, ERWIN AND KNECHT, GERHARDT: Betriebslehre. Eugen Ulmer GmbH & Co, Stuttgart, 1995 REISCH, ERWIN AND ZEDDIES, JÜRGEN: Einführung in die landwirtschaftliche Betriebslehre. Eugen Ulmer GmbH & Co, Stuttgart, 1992 BRANDES, WILHELM AND ODENING, MARTIN: Investition, Finanzierung und Wachstum in der Landwirtschaft. Eugen Ulmer GmbH & Co, Stuttgart, 1992 BREALEY, RICHARD and MYERS, STEWART and FRANKLIN ALLEN: Corporate Finance. McGraw-Hill/Irwin, New York, 2006