Screw-pressing of Jatropha seeds for fuelling

purposes in less developed countries

Eindhoven University of Technology
Department of Sustainable Energy Technology












By: Peter Beerens

First supervisor: prof dr ir A.B. de Haan
Second supervisor: dr ir M.J. Prins
Third supervisor: mr ir ing I.P. Biemond

Advisors: prof ir C. Daey Ouwens
ir P. Willems
dhr R. van Eck

Student number: 0513417
Date: August, 2007
Acknowledgements


i
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Acknowledgements
This thesis was written to graduate for the Master Sustainable Energy Technology at
Eindhoven University of Technology. Occasion to this study was a question from Diligent
Energy Systems on how to optimize Jatropha oil production. The study consisted of
practical research at the university and a two and a half month visit to Tanzania for both
practical research and field studies.

I would like to use this opportunity to thank the people in the Netherlands and Tanzania
who made my graduation largely a pleasant experience. If it had not been for the
enthusiasm of Kees Daey Ouwens I would never have started this research about
Jatropha and the pressing process in particular. I want to thank him for helping me to
shape the project from the start, his feedback during the process and providing me the
opportunity to visit Tanzania. Furthermore I am very glad that André de Haan was willing
to supervise my project with such enthusiasm, thereby enabling me to research the topic
that appealed to me. As I depended on the help of many persons I want to express my
thanks by mentioning them here:
• Ies Biemond and Mark Prins, my second and third supervisor
• Paul Willems, for helping me with the setup of my practical research and the
extensive feedback on this thesis
• Ruud van Eck, for adding practical relevance to my research and enabling me to
visit Diligent Energy Systems in Tanzania.
• Janske van Eijck, for receiving me at Diligent Tanzania and making me feel at
home.
• All employees of Diligent Tanzania LTD, for their warm welcome and their help to
complete my research.
• Supporting staff members of Combustion Technology at the faculty of mechanical
engineering; especially Frank Seegers and Gerard van Hout who helped me with
most practical arrangements.
• Supporting staff members of System Process Engineering at the faculty of
chemical engineering, for helping me with the setup of my analysis methods.
• Niels Ansø, for providing me with spare parts for the screw press free of charge.
• All the people that I forgot to mention above who contributed to this thesis.
• The companies Reinartz and Keller, for providing me with the results of their
Jatropha press tests.

Finally I want to thank my girlfriend, my friends and my parents who gave me the
support, thrust and opportunities to distract my mind in order to complete my
graduation.

Peter Beerens,
Einhoven, July 2007.



i
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Summary


ii
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Summary
People living in the poorest regions in the world, e.g. large parts of the African, Asian and
Latin American continents, often lack access to energy sources in general. One approach
to provide these people with energy to increase living standards is to enable them to
produce energy from local resources. A promising local renewable energy source for
people living in tropical regions is Jatropha Curcas L; a plant that grows oil yielding seeds
that contain 31-38% oil.

The seeds need to be processed in order to obtain the useful end product oil. Mechanical
screw pressing is the most popular method in the world to separate oil from vegetable
oilseeds on small to medium scale. Therefore the impact of several variables on oil
recovery, oil quality and energy costs for screw pressing is reviewed in this report. The
main question to be answered in this report is:

• How can oil recovery, energy input and cost of Jatropha seed expression in screw
presses be optimized while retaining a quality that complies with international fuel
standards?

Additional consideration of the practical application of screw presses for the production of
Jatropha oil in Tanzania led to answers to the following two additional questions.

• What is the optimal setup for the production of Jatropha oil in Tanzania?
• In what way can rural areas in which their use is anticipated benefit from Jatropha
cultivation?

Two distinct low capacity screw presses were used; one located in the Netherland and the
other one in Tanzania. Comparing the results of test on both presses reveals which of the
two types is best suited to extract oil from Jatropha seeds. The following variables were
in advance considered of main interest:

• Size of the restriction at the end of the press chamber
• Rotational speed of the screw press shaft
• Moisture content of the seeds
• Hull content of the seeds

The most important conclusions that can be drawn from this study are as follows:
Moisture content has the strongest effect on oil recovery. Restriction size and rotational
speed of the screw are other influential parameters. The highest oil recovery measured
were 89% and 91% for the BT50 and Sayari press respectively. These values resulted
after one hour cooking in water of 70°C. Oil recovery values for untreated seeds under
standard circumstances were 79% and 87%. Important to note is that the Sayari expeller
requires dual passing of the material compared to single passing for the BT50. Additional
Jatropha press tests were conducted at two industrial German press producers in order to
put the other results into perspective. The tests conducted during this research showed
close resemblance to industrial practice. Taking into consideration all the test results
optimal oil recovery is expected at a moisture content of 2-4% after cooking at 70°C,
100% hull content and the smallest restriction size and lowest speed possible for a
certain press type.

Based on purely financial results from the ‘cost benefit analysis’ large scale centralized
processing is preferred over decentralized pressing. For similar production scale
centralization allows an ‘internal rate of return’ of 61% compared to 30-40% in case of
decentralization. However, reviewing both financial and non-financial aspects shifts
preference to decentralized pressing for Tanzania and similar countries.

Summary


iii
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Jatropha has significant added value for rural communities in Tanzania. For a village of
9000 inhabitants Jatropha cultivation has an added value between $12,750 and $54,500
depending on seed yield, which is equivalent to approximately 100 annual minimum
wages. Oil processing on village scale without the intervention of commercial firms seems
preferable over mere seed selling if coupled to electricity generation and maize milling.
100 acres of Jatropha could easily provide an additional $1,500 per annum, which is
equivalent to 19 monthly salaries to be divided between the initiators.
Table of content


iv
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Table of content

Acknowledgements i
Summary ii
1 Introduction 1
1.1 Occasion 1
1.1.1 Background 1
1.1.2 Problem 1
1.2 Goals 2
1.2.1 Research question 2
1.2.2 Working Method 3
1.2.3 Research limitations 3
1.3 Report structure 4
2 Jatropha and its applications 5
2.1 Plant description 5
2.2 Seed properties 6
2.3 Jatropha applications 7
2.3.1 Fuel 7
2.3.1.1 PPO 8
2.3.1.2 Biodiesel 9
3 The screw press 11
3.1 Working principles 11
3.2 The two main screw press designs 14
3.2.1 Strainer press 14
3.2.2 Cylinder-hole press 14
3.2.3 Different designs 15
3.3 Parameters affecting the screw pressing process for oilseeds 16
3.3.1 Independent variables 16
3.3.2 Dependent variables 16
4 Materials and method 18
4.1 Experiments at Eindhoven University of Technology 18
4.1.1 Materials 18
4.1.2 Experimental setup 18
4.1.3 Experimental procedure 19
4.1.4 Analysis 20
4.1.5 Accuracy and reproducibility of experimental results 21
4.2 Experiments at Diligent Tanzania LTD 21
4.2.1 Materials 21
4.2.2 Experimental setup 21
4.2.3 Experimental procedure 22
4.2.4 Analysis 23
4.2.5 Accuracy and reproducibility of experimental results 24
5 Results and discussion 25
5.1 Data processing 26
5.2 Influence of RPM 27
5.3 Influence of restriction size 28
5.4 Influence of moisture content 30
5.5 Influence of hull fraction 32
5.6 Influence of seed preheating 34
5.7 Influence of other pre-treatments 35
5.8 The effects on foots 35
5.9 Oil quality parameters 36
5.10 Comparison of suitability for Jatropha processing between BT50 and Sayari 36
5.10.1 Test conducted by industrial press manufacturers 36
5.11 Conclusions 37
6 Possibilities for oil pressing in rural Tanzania 39
6.1 Introduction 39
Table of content


v
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
6.2 Cost Benefit Analysis: Theory 39
6.2.1 Profitability assessment 40
6.2.2 Liquidity assessment 41
6.2.3 Sensitivity analysis 41
6.3 CBA results: comparison of centralized and decentralized production 41
6.3.1 Decentralized pressing 43
6.3.2 Centralized pressing 44
6.3.3 Sensitivity analysis 45
6.3.4 Non financial consideration 45
6.4 Viability of community level processing: Engaruka case 46
6.5 Conclusions 48
7 Conclusions and recommendations 50
Reference List 52
Appendix A: Properties of Jatropha seeds and oil 56
Appendix B: German quality standard for Pure Plant Oil 57
Appendix C: Data on BT 50 bio press setup 58
Appendix D: Maximum power output and thermocouple calibration 59
Appendix E: Measurement equipment in Tanzania 60
Appendix F: Sayari expeller and Intermech workshop 62
Appendix G: Overview of variables included in this research 63
Appendix H: Calculation of oil recovery from soxhlet extraction 64
Appendix I: Initial measurement data 65
Appendix J: Calculation of energy requirement 67
Appendix K: Foot levels in oil from the Sayari expeller and BT50 press 70
Appendix L: Reinartz & Keller presses and industrial oil recovery percentages 72
Appendix M: CBA results for different process setup 75
Appendix N: Calculation of the Engaruka case 77
Appendix O: Villages for case study 78
Chapter 1: Introduction


1
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
1 Introduction
This chapter provides some background information important for this research.
Furthermore the research problem and goals are stated. Working method and research
limitations are also shortly elaborated on. Finally the structure of this report is shortly
described.
1.1 Occasion
1.1.1 Background
People living in the poorest regions in the world, e.g. large parts of the African, Asian and
Latin American continents, often lack access to energy sources in general. For cooking
they often depend on gathered fuel wood and in case a connection to the electricity grid
is present it is often unreliable. One approach to provide these people with energy to
increase living standards is to enable them to produce energy from local resources. By
means of such an approach the reliance on help from foreign countries is limited, which
could lead to long term improvement of well-being. A promising local renewable energy
source for people living in tropical regions is Jatropha Curcas L; a plant that grows oil
yielding seeds. What distinguishes Jatropa from many other bio fuel crops are the
benefits it can offer to relatively small rural areas in Less Developed Countries. Some
examples are use for fencing, decreased deforestation for firewood collection and
reduction of soil erosion. Once the oil is removed from the seeds, either mechanically or
chemically, the oil and its by-products can serve various purposes some of which are:
liquid fuel, gas, electricity, fertilizer and soap (Openshaw, 2000, 1-19).

Mechanical screw pressing is the most popular method in the world to separate oil from
vegetable oilseeds on small to medium scale. Reasons for its popularity in for example
India are that the machines require low initial and operation investments, can easily be
operated, maintained and adopted by semi-skilled personnel (Bargale and Singh, 2000,
130-134). Furthermore it is possible to manufacture screw presses locally creating
additional local employment. Both the oil and the de-oiled press cake obtained using
screw presses are free of solvents and other chemicals as opposed to the more efficient
solvent extraction method.

Research on mechanical screw presses for pressing oilseeds dates back to 1951 when
V.D. Anderson Company patented the first expeller. Ever since considerable efforts have
been made to gain better understanding of the processes inside a screw press and
improve the oil extraction efficiency (Bargale and Singh, 2000, 130-134;Vadke and
Sosulski, 1988, 1169-1176;Ward, 1976, 261-264;Zheng and others, 2003, 1039-
1045;Zheng and others, 2005, 193-202). The earlier research was aimed at optimizing
process variables like pressure, temperature and rotational speed of the press worm. In
addition changes in moisture content and physical, thermal and hydrothermal pre-
treatment of the seeds were considered in later research. The effect of research
outcomes was that oil recovery rates increased from 50% to 80% for various oilseeds
(Bargale and Singh, 2000, 130-134). Based on previous research various pre-treatments
and process variables are considered for the present study, taking into consideration the
possible application in rural Africa.

1.1.2 Problem
Although wild varieties of Jatropha exist in many regions attempts to cultivate the plant
are limited. Cultivation experiments have been carried out in the previous decade and
actual plantations have emerged in Africa, India and Latin America over the last few
years. To name a few countries: Belize, Nicaragua, Ghana, Tanzania, Zambia, Mali and
Mozambique (Hartlieb Euler and David Gorriz, 2004;Henning, 2007). There is however a
Chapter 1: Introduction


2
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
lot of uncertainty on the best way to process the seeds into oil, taking into account the
backward circumstances in less developed countries.

One of the disadvantages of screw pressing is the low oil recovery when processing
untreated seeds. Even after multiple passes 5% to 10% of the oil remains in the press
cake after pressing (Shahidi, 2005). Previous studies showed that oil recovery rates can
be raised from 73% to 80% of the initial oil content for rapeseed and peanut and from
60% to 65% for cotton seeds by improvement of the press settings and proper the
economic viability of local oil production by screw pressing and should therefore also be
examined for Jatropha.
1.2 Goals
1.2.1 Research question
The most interesting application for the Jatropha oil is to use it as a fuel. This way the oil
can be used for mobility (vehicles), electricity (generator), lighting and cooking. Once the
intended use of the product is known, the production process can be chosen and
analysed. The three most important aspects to consider when analysing expeller
operation are the amount of oil removed, the amount of energy required to achieve this
and the quality of the oil with respect to the intended application. Therefore these are
also the aspects touched upon in the main research question:

• How can oil recovery, energy input and cost of Jatropha seed expression in screw
presses be optimized while retaining a quality that complies with international fuel
standards?

In this question oil recovery is defined as the amount of oil removed from the seeds compared
to the amount that was initially present in the seeds.

There are three possible approaches to improve the properties mentioned in the research
question namely pre-treatment of seeds before pressing, changes in press operation and
changes in press design. However this research is extended beyond the determination of
the optimal condition under which Jatropha seeds should be processed. The success of a
new technology or production process is inseparably linked with the way it is introduced.
It might therefore be worth the effort to go in search of an appropriate implementation
approach for each specific project. For this research the north-eastern region of Tanzania
was selected as a case study. In and around the cities of Arusha and Moshi attempts are
made to set up a complete system to grow, process and use Jatropha to its full potential.
Diligent Energy Systems is a Dutch company operating in the field of bio fuel production
in less developed countries and active in Tanzania and Colombia. In 2003 Diligent started
to contact local farmers and create networks for the collection of Jatropha seeds. By the
end of 2006 their supply of seeds became sufficient to go on to the next step, which is
processing the seeds into high quality oil. The problems Diligent encountered in
determining the best suited processing method for Jatropha seeds in Tanzania gave
occasion to this research. A moment of reflection reveals that this challenge not only
requires a technical solution, but also demands investigation of local circumstances.
Numerous examples exist of innovations that where technically sound, but still failed
dramatically (Douthwaite, 2002). Therefore a second and third question are included to
identify and avoid some of the pitfalls correlated with the introduction of innovations in
less developed countries:

• What is the optimal setup for the production of Jatropha oil in Tanzania?

• In what way can rural areas in which their use is anticipated benefit from Jatropha
cultivation?
Chapter 1: Introduction


3
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
1.2.2 Working Method
Two distinct low capacity screw presses were used; one located in the Netherland and the
other one in Tanzania. Comparing the results of test on both presses reveals which of the
two types is best suited to extract oil from Jatropha seeds. Both screw press settings and
properties of the Jatropha seeds are varied to get a better understanding of the
expression process and to find the optimal process conditions. The extent to which oil is
separated from solid material is determined by analysing and comparing the oil content
of both input (seeds) and output material (press cake). Based on literature the most
important variables influencing the process will be (Bargale and Singh, 2000, 130-
134;Eggers, Broeck, and Stein, 2006, 494-499;Vadke and Sosulski, 1988, 1169-
1176;Zheng and others, 2003, 1039-1045):

• Size of the restriction at the end of the press chamber
• Rotational speed of the screw press shaft
• Moisture content of the seeds
• Hull content of the seeds

Because Jatropha oil is intended to be used in diesel engines it is important to compare
the quality of the oil according to international standard for pure plant oils (PPO) and
biodiesel. As this involves many different and time consuming chemical analyses, testing
all fuel quality parameters is outside the scope of this graduation project. Indispensable
tests on for example acidity and phosphorous content were contracted out to external
research institutes.

During a visit to Tanzania the test results gathered at Eindhoven University of Technology
were compared to a different type of screw press operating at Diligent’s workshop in
Arusha, Tanzania. The main goal of the visit was to identify local circumstances and
technological means to design a plan for local expression of Jatropha oil. At the moment
of writing there is little activity involved in organized pressing of the seeds. Field studies
were conducted from Jan-Mar 2007 consisting of visits to farmers who grow Jatropha, a
local producer of screw presses, running Jatropha presses, involved Non government
Organisations (NGOs), universities actively researching Jatropha and other interesting
opportunities that presented itself. The collected data was used to perform a ‘Cost Benefit
Analysis’ (CBA) on optimal Jatropha processing scale in Tanzania.
1.2.3 Research limitations
In this report the impact of several process variables on oil recovery, oil quality and
energy costs are reviewed. The results are obtained by practical tests on two different
standard types of screw presses; namely cylinder-whole press and strainer press. The
practical research is restricted to variation of pre-treatments and press operation.
Changes in screw design in order to improve oil yield were not included in this study. In
addition costs and maintenance required for operation are estimated.

The only method of oil extraction examined in the report is mechanical extraction.
Solvent extraction can achieve a reduction in the oil content of the press cake to less
than 1% (Shahidi, 2005). In spite of its higher oil yields solvent extraction has some
drawbacks especially for small scale applications. Solvent extraction facilities are typically
constructed to process several thousands of tons of seed per day and cost in the order of
tens of million dollars (Shahidi, 2005). In addition the commercially used solvent hexane
can become highly explosive when mixed with air. Pre-treatments for size reduction and
cell rupturing are also required prior to solvent extraction adding extra costs. For the
purpose of pressing oilseeds on a small scale in less developed countries screw presses
seem to be the better option despite their lower yield. As current methods of solvent
extraction are not applicable in the region of interest they are not treated in further detail
in this study.

Chapter 1: Introduction


4
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Further treatment of the oil that comes out of the press might be required before it can
be used to properly run a diesel engine. Although these steps are touched upon in the
report no tests were carried out on the effect of different treatments of the oil on the
performance or emission of a diesel engine.

The examination of socio-technological aspect in addition to the technical aspects is
considered an added value to this report. Presumably the performed CBA will be less
thorough than would be the case for a graduation project fully dedicated to this subject.
1.3 Report structure
The Jatropha plant and the applications of the oil and its by-products are elaborated on in
more detail in chapter 2. In chapter 3 the working principles of different types of screw
presses are explained. In this third chapter the variables affecting the pressing process
are also discussed. Chapter 4 focuses on the research method of the experiments that
were carried out on the screw press and the chemical analyses used to analyse the press
cake. The research results are explained and discussed in chapter 5. Chapter 6 provides
an overview of the findings in Tanzania. Based on the identification of local circumstances
and a Cost Benefit Analyses (CBA), ideas on how to setup Jatropha pressing in Tanzania
is discussed. Finally conclusions and recommendations are given in chapter 7.
Chapter 2: Jatropha and its applications


5
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
2 Jatropha and its applications
This chapter serves as both an introduction to Jatropha for the layman and a detailed
description of characteristics of Jatropha seeds and oil related to screw pressing.
Paragraph 2.1 is a general description of the Jatropha tree. In paragraph 2.2 some
interesting seed properties with regards to screw pressing are mentioned. The third
paragraph deals with the various applications for Jatropha products and specifically Pure
Plant Oil and biodiesel.
2.1 Plant description
When people use the term Jatropha, usually they refer to the species Jatropha Curcas L.,
which is one of the 170 known species of this plant (2006;Paramathma and others,
2006). The species Jatropha Curcas, of particular interest for this study, will for
simplicities sake from now on be referred to as Jatropha. Jatropha is a wild plant; in
other words it has not been cultivated through variety research. The plant belongs to the
family Euphorbiaceae and is indigenous to Latin America (Akintayo, 2004, 307-310) and
naturalized throughout tropical and subtropical parts of Asia and Africa where it is mostly
used for hedging (Augustus, Jayabalan, and Seiler, 2002, 161-164). Both plant and seeds
are toxic to humans and animals due to the presence of phorbol esters (Haas and
Mittelbach, 2000, 111-118). Seeds from the Jatropha tree are known by many different
names, the most common of which is “physic nut”. In Tanzania it is known as “Mbono” or
“Makaen” and in the Netherlands as “purgeernoot”. Jatopha is a bush or small tree that is
able to survive on marginal lands and can get up to 6 or 8 meters high. Its leaves have
length and width of 6 to 15 cm and are shaped as is visible in Figure 2-1. This perennial


Figure 2-1 Left: Representation of the Jatropha plant (a,b,c,d), fruits (g,h) and seeds (i) Right:
close-up of Jatropha fruits near Arusha, Tanzania (Heller, 1996).

bush can live up to 50 years and starts to grow fruits from the 2
nd
or 3
rd
year. Under
favourable conditions it can grow to a thick bushy fence of approximately one meter high
in 6-9 months (Augustus, Jayabalan, and Seiler, 2002, 161-164). Jatropha will grow
under a wide range of rainfall regimes from 200-1500mm. Ideal would be 600mm in
moderate climates and 1200mm in hot climates (Biswas, Kaushik, and Srikanth, 2006).
Being drought tolerant the plant still sheds its leafs during dry periods, which means that
it cannot grow fruits (Openshaw, 2000, 1-19). In addition to water a plant needs

Chapter 2: Jatropha and its applications


6
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
nutrients. In case of Jatropha deficiency of these nutrients leads to increased failure of
seed development as was encountered during my field studies in Tanzania in February
2007. The plants yield varies strongly and although it survives on marginal land, yields
will be significantly lower than for fertile lands. Claims on yields range from 0.4 to 12
ton/hectare/year (Openshaw, 2000, 1-19). The number of trees per hectare has been
reported to vary from 1100-3300, which equals 445-1335 per acre (Openshaw, 2000, 1-
19). The seed yield of Jatropha has been estimated to range from 0.75-2 kg/year per
plant after three to four years (Biswas, Kaushik, and Srikanth, 2006). In contrast to the
overoptimistic claims of 12 tons/hectare (4.8 tons/acre) calculation using the above
values predict yield between 0.8-6.6 tons/hectare (0.33-2.7 tons/acre). This seems a
reasonable when compared to claims in the comparative studies by Benge (Benge,
2006).
2.2 Seed properties
Much research has been conducted on the composition and properties of Jatropha seeds
by others ((Openshaw, 2000, 1-19), (Heller, 1996), (Henning, 2004) and (Sirisomboom,
7 A.D.)). These studies also provide insight in the possibilities of using Jatropha oil for
fuel purposes. Knowledge of the physical and mechanical properties of Jatropha is
required for adequate design of machines for de-hulling, drying and pressing of the
seeds. Restrictions on how to store the seeds are linked to these properties. Mechanical
properties such as rupture force and energy required for rupturing fruit, nut and kernel
provide insights on how to adapt the pressing process to Jatropha seeds.

Table 2-1 Physical and mechanical properties of Jatropha (Sirisomboom, 7 A.D.). Note that the data
concerns fresh seeds and that under normal conditions the density is less than that of water.
Physical properties Nut Kernel
Length [mm] 21.02 ± 1.03 15.45 ± 0.54
Equatorial width perpendicular to length [mm] 9.58 ± 0.28 7.42 ± 0.33
Breadth perpendicular to length and width [mm] 11.97 ± 0.30 10.25 ± 0.36
Solid density [kg/m
3
] 1040 1020
Bulk density [kg/m
3
]
450 420
Mechanical properties Nut Kernel
Rupture force [N] 146.63 ± 14.82 67.72 ± 19.03
Hardness [N/mm] 69.98 ± 6.22 38.52 ± 5.59
Energy used for rupture [Nmm] 124.44 ± 19.95 51.61 ± 26.84


Jatropha fruits are slightly elliptical in shape with length approximately 35 mm and
contain 3 seeds on average (Sirisomboom, 7 A.D.). The fruits and seeds are shown in
Figure 2-1. The size of the seeds varies between 11-30 mm (average 21mm) in length
and 7-12 mm (average 11mm) in width (2006;Sirisomboom, 7 A.D.). More detailed
information on the physical and mechanical properties of Jatropha seeds is given in Table
2-1. Hardness values in Table 2-1 indicate that Jatropha seeds are relatively soft
compared to for example rapeseed (>52.6 N/mm) (Faborode and Favier, 1996, 335-345)
and sunflower (35.3-65.3 depending on seed orientation) (Gupta and Das, 2000, 1-8).
The seeds weigh about 1 ton/m
3
. The oil content of the seeds varies with origin and
growing conditions and is between 30-40 wt.%, which makes it a high oil content seed.
Variations of oil content with origin are shown in Table 2-2. For Tanzanian varieties this is
approximately 38%. The influence of seed quality on oil content could not be checked as
the samples were provided by various parties.

Chapter 2: Jatropha and its applications


7
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Table 2-2 Oil content of seeds of different origin determined by soxhlet extraction
Origin Oil content
Brasil 30.9%
Tanzania 37.8%
Ethiopia 38.8%
India 36.8%
Gambia 32.7%
Nigeria 33.7%

2.3 Jatropha applications
Jatropha can be utilized for various purposes of which application as transport fuel is
probably the most interesting one from both an economical and ecological point of view.
Nevertheless the other uses are worth mentioning as they provide insight in the total
value chain of Jatropha products. In rural Africa Jatropha is mostly used as a natural
fence to prevent cattle from feeding on crops for medicinal purposes and more recently
for soap making by local women groups. The bark and leaves contain dye and latex with
medicinal properties for the pharmaceutical industry (Augustus, Jayabalan, and Seiler,
2002, 161-164). As mentioned earlier one drawback of Jatropha is that there are hardly
any yields during the first two or three years. One option to overcome this problem is
intercropping of Jatropha with other crops like for example maize. In Madagascar
Jatropha is, for the same reason, used to support vanilla plants (Benge, 2006). If planted
wisely the plant can be used to prevent soil erosion or even reclaim exhausted land.
When managed properly it has potential as a commercial crop for the production of Pure
Plant Oil or biodiesel. The press cake formed during oil production also has a wide variety
of applications depending on local circumstances. As it still contains most of the nutrients
it can be used to fertilize land. Other options are the production of briquettes for
industrial steam boilers or charcoal for household stoves. A GTZ (Gesellshaft für
Technische Zusammenarbeit) project with plant oil stoves in Tanzania indicates that the
oil might also have some potential as fuel for cooking stoves. Possibilities for biogas
production from Jatropha press cake have also been studied (2006;Aderibigbe and
others, 1997, 223-243).
2.3.1 Fuel
In spite of all other options fuel is considered the most promising application for the
Jatropha seeds. The potential work that a fuel can do is determined by its energy
content. In
Table 2-3 the calorific values of solid and liquid Jatropha fuels are compared to those of
conventional fuels. More detailed information on seed and oil properties is provided in
Appendix A. An interesting value when judging a fuel is the energy value. The numbers in
Table 2-3 suggest that, judging on energy value, liquid fuel is the highest-grade product
to be obtained from Jatropha seeds. In addition press cake seems a valuable by-product.
One should think of using the press cake for the production of for example briquettes,
charcoal or biogas (Vyas and Singh, 2007, 512-517). Jatropha press cake has an energy
content between that of coal and wood. Jatropha oil shows lower energy content than
conventional fuels. This indicates higher fuel consumption per kilometre. When
converting to the more interesting quantity MJ/litre the values for Jatropha and diesel
change to 34.77 MJ/litre and 37.72 MJ/litre respectively, reducing the difference in
energy content to 8%. Most Jatropha related activities as well as this report are centred
on liquid fuel. Therefore PPO and biodiesel from Jatropha are discussed in more detail in
paragraph 2.3.1.1 and 2.3.1.2.
Chapter 2: Jatropha and its applications


8
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

Table 2-3 Calorific value of Jatropha oil and seeds compared to conventional fuels (Augustus,
Jayabalan, and Seiler, 2002, 161-164).
Parameter Calorific value (MJ/kg)
Jatropha seeds 20.1 ± 0.08
Jatropha press cake 18-25.1**
General purpose coal
(5–10% water)
*
32–42
Wood (15% water)
*
16
Jatropha oil 37.8 ± 0.08
Biomass and fossil fuels
Fuel oil (mexico) 43.158
Crude oil 44.091
Gasoline 47.127
Diesel
*
46
* http://www.kayelaby.npl.co.uk
** depending on residual oil content in the press cake

2.3.1.1 PPO
Vegetable oil that has not been treated apart from filtering is often referred to as Pure
Plant oil (PPO) or Straight Vegetable Oil (SVO). Jatropha oil is just one of a wide variety
of PPOs. The main reason to use PPO instead of converting it to biodiesel is the costs
involved in the transesterification process. Compared to conventional diesel the use of
PPO in a diesel engine reduces the emission of sulphur oxides, carbon monoxides, poly
aromatic hydrocarbons, smoke, particle matter and noise (Agarwal and Agarwal, 2007,
2314-2323). The main disadvantage of PPO on the other hand is its high viscosity that
leads to unsuitable pumping and fuel spray characteristics. The high viscosity of Jatropha
oil is a result of the presence of the saturated and unsaturated acids identified in Table
2-4. The acids consist of relatively long carbon chains (C14:0-C20:0) when compared to
conventional diesel (C12:0-C15:0) (Knothe and Steidley, 2005, 1059-1065). The acid
composition in Table 2-4 also hints an acidic nature of Jatropha oil, which is harmful to
rubber engine components. Poor atomization, low volatility and insufficient fuel-air
mixing can lead to combustion chamber deposits, gum formation and unburned fuel in
the lubrication oil (Agarwal and Agarwal, 2007, 2314-2323).

Table 2-4 Jatropha oil fatty acid composition (Akintayo, 2004, 307-310;Augustus, Jayabalan, and
Seiler, 2002, 161-164;Kandpal and Madan, 1995, 159-160;Mohibbe Azam, Waris, and Nahar, 2005,
293-302))
Saturated acid Augustus,2002 Akintayo, 2003 Kandpal & Madan, 1994
Palmatic acid (16:0) 14.1% 19.5% ± 0.8 12.8%
Stearic acid (18:0) 6.7% 6.8% ± 0.6 7.3%
Unsaturated acid
Oleic acid (18:1) 47.0% 41.3% ± 1.5 44.8%
Linoleic acid (18:2) 31.6% 31.4% ± 1.2 34.0%


In Table 2-5 a comparison between Jatropha and conventional diesel is made. The
enormous difference in viscosity is striking. The high saponification together with a low
unsaponifiable mass shown in Appendix A indicates that Jatropha oil is a normal
triglyceride. Although much higher than the viscosity of normal diesel or biodiesel,
standards of which are between 1.6 and 6 cST, the viscosity of Jatropha (34-36cST) is
average compared to other plant oils such as soybean (31cST), cottonseed (36cST) and
sunflower (43cST) (Knothe and Steidley, 2005, 1059-1065). Its viscosity being
Chapter 2: Jatropha and its applications


9
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
comparable to other plant oils makes Jatropha a potential material for fuel production,
although measures to reduce it seem necessary (Akintayo, 2004, 307-310). German PPO
norms are included in Appendix B to show the other specifications to comply with.

Table 2-5 Some fuel properties of Jatropha oil and mineral diesel compared (Agarwal and Agarwal,
2007, 2314-2323).
Parameter Fuel
Mineral diesel Jatropha oil
Density (kg/m
3
) 840 ± 1.732 917 ± 1
Kinematic viscosity at 40°C (cST) 2.44 ± 0.27 35.98 ± 1.3
Flash point (°C) 71 ± 3 229 ± 4
Fire point (°C) 103 ± 3 274 ± 3
Calorific value (MJ/kg) 45.343 39.071
Carbon (wt%) 80.33 76.11
Hydrogen (wt%) 12.36 10.52
Nitrogen (wt%) 1.76 0
Oxygen (wt%) 1.19 11.06
Sulfer (wt%) 0.25 0

2.3.1.2 Biodiesel
Biodiesel is a fuel type that has already been proven suitable for use in diesel engines.
Biodiesel consists of fatty acid methyl esters (FAMEs) of seed oil and fats and is produced
through transesterification. The reaction equation for biodiesel production is shown in
Figure 2-2. The triglyceride esters of oil are changed into methanol monoesters (methyl
esters), each with single fatty acid chains causing the lower viscosity of biodiesel. FAMEs
are environmentally safe, non-toxic and biodegradable making them a suitable transport
fuel (Mohibbe Azam, Waris, and Nahar, 2005, 293-302). The biggest advantage of
biodiesel over PPO is its lower viscosity (Knothe and Steidley, 2005, 1059-1065).


Figure 2-2 Reaction equation of transesterification (www.biofuel.be)

One of the important values for selection of FAMEs for biodiesel is the cetane number
(CN). This is the ability of a fuel to ignite quickly after injection. Higher values indicate
better ignition quality. Some of the most widely used biodiesel standards from the USA
(ASTM D 6751), Germany (DIN 51606) and European Organization (EN 14214) set CN at
47, 49 and 51 respectively (Mohibbe Azam, Waris, and Nahar, 2005, 293-302). The CN
for Jatropha biodiesel is on the high end being 52.31 (Mohibbe Azam, Waris, and Nahar,
2005, 293-302).

Another important criterion for biodiesel quality is the iodine value (IV), which indicates
the degree of unsaturation (amount of double bonds). Some level of unsaturated fatty
acid components is necessary to prevent FAMEs from solidification. However at too high
IV unsaturated molecules start to react with atmospheric oxygen and convert to peroxide
which might lead to the polymerization of biodiesel into plastic-like material. High
temperatures occurring in internal combustion engines tend to accelerate this process. To
comply with all standards the IV should be less than 115 (Mohibbe Azam, Waris, and
Chapter 2: Jatropha and its applications


10
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Nahar, 2005, 293-302). With an IV of 93.0 (Mohibbe Azam, Waris, and Nahar, 2005,
293-302) Jatropha biodiesel is well within range.

Research by Mohibbe Amam et al, 2005 points out Jatropha biodiesel to be among the
twenty-six best suited FAME mixtures (from a 75 species comparison) judging CN, IV and
other variables like linolenic acid value, boiling point and carbon chain length. The latter
causing higher viscosity with increased chain length (Krisnangkura, Yimsuwan, and
Pairintra, 2006, 107-113).
Chapter 3: The screw press


11
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
3 The screw press
Chapter 3 is intended to give the reader more information about the screw pressing
process. Therefore paragraph 3.1 first explains the working principles of oil expelling in a
screw press. The two main screw press designs are treated in paragraph 3.2. Finally
paragraph 3.3 goes into more detail on the parameters affecting the screw pressing
process.
3.1 Working principles
Mechanical pressing and solvent extraction are the most commonly used methods for
commercial oil extraction. Screw pressing is used for oil recovery up to 90-95%, while
solvent extraction is capable of extracting 99% (Shahidi, 2005). In spite of its slightly
lower yield, screw pressing is the most popular oil extraction method as the process is
simple, continues, flexible and safe (Zheng and others, 2005, 193-202) (Singh and
Bargale, 2000, 75-82). Screw presses exist with raw seed capacities ranging from 10 kg
/hr up to tens of tons/hr. The operation principle of a screw press is rather simple to
visualize although very difficult to model judging from attempts by Shirato (1971), Vadke
(1988) and Wang (2004). Therefore most improvements made to screw presses over the
last decades are based on manufacturer’s experiences and intuition rather than on the
basis of physical principles (Vadke and Sosulski, 1988, 1169-1176); (Singh and Bargale,
2000, 75-82). During the pressing process oil seeds are fed in a hopper and then
transported and crushed by a rotating screw in the direction of a restriction (sometimes
referred to as die or nozzle). As the feeding section of a screw press is loosely filled with
seed material the first step of the compression process consists of rolling, breaking and
the displacement and removal of air from intermaterial voids. As soon as the voids
diminish the seeds start to resist the applied force through mutual contact (Faborode and
Favier, 1996, 335-345). The continuous transport of material causes pressure to increase
to a level needed to overcome the restriction. This pressure causes the oil to be
expressed from the seeds. The end pressures in a screw press can vary from 40 to 350
bar (4-35 MPA) depending on the type of press and input material.

Material conveyance through the press barrel is a result of the friction forces between
material, screw and barrel. This force balance is represented in simplified version in
Figure 3-1. If F
p
(induced by the screw) and F
r4
(friction between material and barrel) are
larger than the sum of the other friction and pressure forces the material will be properly
transported. When the friction forces between screw and seed material increase beyond
F
r4
slip will occur and transport will reduce or even stop.

The oil is situated at different locations inside the cell, together with other constituents
like proteins, globoids and nucleus represented in Figure 3-2. All these elements are
contained within cell walls, which need to be ruptured to free the oil. A combined force of
friction and pressure in the barrel causes the cell walls to rupture and oil to flow out of
the liquid solid mixture inside the barrel. The separated oil is discharged through holes
provided along the press barrel. The compressed solid material or press cake is
simultaneously discharged through the restriction at the end of the barrel. The various
screw press components have been described in detail by Ward (1976), Singh & Bargale
(2000) and Shukla (1990).

Chapter 3: The screw press


12
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

Figure 3-1 Forces acting on a press cake element situated between two worm shafts. The x-
direction indicates movement towards the restriction. (Eggers, Broeck, and Stein, 2006, 494-499)

Expression of liquid solid mixtures generally consists of two steps: filtration and
consolidation. Whenever free flowing particles are present the process is called filtration.
From the point where all solids combine into a solid bed it is called consolidation
(Willems, 2007). The extent to which consolidation occurs is mainly determined by the
geometry of the extruder channel and outlet and material properties. Filtration is
expected not to be present inside a screw press (Venter and others, 2006, 350-358) and
therefore consolidation is the main process of interest here. Further distinction can be
made between primary consolidation comprising rearrangement of the solid bed and
liquid removal, and secondary consolidation consisting of material deformation and
further liquid removal (Willems, 2007).


Figure 3-2 Cell composition of oilseeds, in this case rapeseed (Singh and Bargale, 1990, 106-110).

Modelling attempts
Extrusion of polymers, which closely resembles oil expelling, is more easily understood
because of a continuous material flow in and out of the press chamber. All that comes in
at the entrance comes out at a single exit, which makes drawing up mass and energy
balances relatively easy. Oil expelling is more complicated as a radial oil flow leaves the
press barrel at the oil outlet, thus creating a second outflow of material as is illustrated in
Figure 3-3 (Vadke, Sosulski, and Shook, 1988, 1610-1616). This oil flow is difficult to
model and therefore theoretical models have not yet been used to improve expeller
design. In the 1970’s and 1980’s the flow of liquid from a solid-liquid mixture has been
studied based on Terzaghi’s (1943) consolidation theory. Mrema and McNulty (1985)
used consolidation theory to predict oil flow from rapeseed and cashew nuts in a piston
ram-press. Furthermore Körmendy (1974) and Shirato (1971) developed mathematical
models for the mechanical separation of liquid from clay-water substances. In the late
1970’s Shirato et al extended their analysis towards continuous expression in screw
Chapter 3: The screw press


13
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
presses. Their model for the prediction of the amount of water separated from the clay-
water mixture still required prior knowledge on axial flow rate of the mixture and the
pressure profile along the worm. This restriction makes the Shirato model unsuited for
design purposes as these variables are not known in advance. Vadke et al (1988)
produced a more complete model to predict the throughput and the reduction in liquid
content in a liquid-solid mixture in a press of unrestricted geometry. Unrestricted
geometry implies that the size of the cylinder and worm can be varied freely in the
model. The most recent attempt to model screw pressing of oilseeds was undertaken by
Willems (2007) and based on the Vadke model including consolidation theory.

Figure 3-3 Material flow in a section of a worm channel during oil expelling (Vadke, Sosulski, and
Shook, 1988, 1610-1616).

Towards the press cake outlet the radial pressure on the mixture inside the barrel
increases due to volumetric compression. The compression is induced by the worm shaft
as it forces more and more material inside the same volume. According to Ward (1976),
who conducted early research on screw presses the pressure distribution along the shaft
would look like Figure 3-4. Work by others shows that the distribution might be shaped
differently as presented in Figure 3-5 (Eggers, Broeck, and Stein, 2006, 494-499).
Recent studies by Willems (2007) suggest that the calculated profile measured by Eggers
is more sensible than the one by Ward. Increased radial pressure increases the outflow of
oil from the cells (Singh & Bargale, 2000). Results by Eggers and Ward were both
intended for the ‘strainer press’ type that will be further explained in the following
paragraph.

Figure 3-4 Pressure variation based on theory of Ward for strainer press (1976)

Chapter 3: The screw press


14
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

Figure 3-5 Pressure distribution along the screw for strainer press (Eggers, Broeck, and Stein, 2006,
494-499)
3.2 The two main screw press designs
Screw presses can be subdivided into two main types, namely the ‘strainer press’ and the
‘cylinder-hole press’. They mainly differ in screw geometry, oil outlet and press cake
restriction.
3.2.1 Strainer press
For this first type the screw rotates in a cage lined with hardened steel bars, resembling
a strainer. Spacers placed between the steel bars permit oil outlet as the pressure on the
feed material increases. The gaps between the bars form an adjustable oil outlet. This
adjustability allows the pressure to be optimized for different input materials. Figure 3-6
shows a schematic of the ‘strainer press’. The figure shows the increasing diameter of the
screw in order to increase pressure in the direction of the choke. Instead of increasing
the screw diameter inserting cones on a constant diameter screw is a frequently applied
alternative. The screw design causes the volume displacement at the feed end to be
considerably larger than at the discharge end. The resulting pressure increase forces the
oil through the strainer (Khan and Hanna, 1983, 495-503). For flexibility often different
screw types are available or subsections can be replaced. The cake is pressed out of the
adjustable choke in the form of flat and often fractured flakes. The choke gap can be
adjusted to influence pressure and the shape and oil content of the press cake. For
smaller machines, a hand wheel on the opposite end of the screw is typically used to
adjust the choke (Khan and Hanna, 1983, 495-503). ‘Strainer presses’ are available with
capacities ranging between 15 and 2000 kg/hour (Ferchau, 2000).

Figure 3-6 Strainer press (Ferchau, 2000).
3.2.2 Cylinder-hole press
With the second press type the oil is pushed out through holes drilled in the cylinder
tube. The profile of this ‘hole cylinder press’ is shown in Figure 3-7. Increasing pressure
forces the press cake through a circular nozzle at the end of the cylinder. In order to
Chapter 3: The screw press


15
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
avoid blocking of the press, the area around the nozzle (press head) is usually preheated
before operation which decreases the viscosity of the paste inside the press. When
pressing rapeseed the temperature of the oil should not increase above 40°C, because
increasing temperature leads to higher phosphor content in the extruded oil. High
phosphor content is an undesirable fuel property as it can cause deposits and clogging in
engines. To prevent the oil from reaching 40°C the temperature near the die, which is the
location where maximum temperature occurs, should at maximum be in the range of 60
to 80°C. If on the other hand the cake outlet temperature is too low the solid content in
the oil becomes unacceptable (Ferchau, 2000).


Figure 3-7 Hole cylinder press (Ferchau, 2000).

The pressure level in the screw press is influenced by nozzle diameter, screw design and
seed conditions. Pressure increases with decreasing nozzle size. This effect would suggest
a small nozzle, in order to extract as much oil as possible. Decreasing the nozzle also has
some negative affects. First of all the seed throughput decreases for smaller nozzle
diameters. Second, due to increased resistance the temperature inside the screw press
increases. This could cause the temperature of the oil to exceed threshold values or lead
to faster machine wear. Thirdly the solid content in the separated oil is higher at higher
pressure caused by reducing the restriction size (Ferchau, 2000). Finally the chance of
blockage increases with decreasing nozzle size. Comparable adaptations are possible for
the strainer based screw press. As a rule of thumb the minimal oil content of the press
cake when using screw presses should be about 5-6% (Hui, 1996). In case of rapeseed a
rotation speed of 20 to 50 rpm is normally used because energy use is low in that case.
3.2.3 Different designs
In addition to conventional expeller design several variations have emerged over time.
One of these variations is a ‘twin-screw extruder’ which is claimed to achieve very high
yields on de-hulled seeds compared to traditional single-screw presses (Dufaure and
others, 1999, 1073-1080). Further advantages are flaking and boiling pre-treatments
would be redundant for ‘twin-screw extruders’. As no literature on the performance of
‘twin-screw extruders’ with Jatropha seeds was found they are mentioned here only for
the sake of completeness and will not be treated in any more detail in this report.
Another variation is solvent injection in an extruder. Claims by Foidl indicate that 0.3%
residual oil content can be achieved when injecting isopropanol under 300 bar pressure
(Foidl, 2007). Foidl also expects 1.5% residual oil content to be possible when using CO
2

as a solvent (Foidl, 2007). Unfortunately no official results to underpin the claims by Foidl
were obtained.
Chapter 3: The screw press


16
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
3.3 Parameters affecting the screw pressing process for
oilseeds
A number of variables need to be taken into account to optimize oil production using
screw presses. For quality preservation temperature is an important parameter. The
friction inside the barrel generates heat which is passed on to the oil. For oil recovery and
energy consumption pressure is more interesting to monitor. A short explanation of the
variables considered in this study is written down below. For more detail on oilseed
processing volume 5 of ‘Bailey’s industrial oil & fat products’ should be consulted.
3.3.1 Independent variables
RPM
Higher screw speed means more throughput and higher residual oil content in the press
cake since less time is available for the oil to drain from the solids. At higher speed the
viscosity thus remains lower resulting in less pressure build-up. This again causes the
residual oil content to be relatively high (Willems, 2007).

Restriction size
When the restriction size is reduced the pressure required to overcome the restriction
increases. A resulting decrease in oil content causes increased viscosity of the paste and
further pressure rise.

Hull content
Hull content is expected to affect both oil recovery and energy requirement. By removing
a part of the hard hull material less energy is needed for breaking and compressing
(Zheng and others, 2005, 193-202). As Jatropha hull contains no oil reduced amounts of
hull may lead to decreased absorption of oil, thereby increasing oil recovery.

Moisture content
An optimal moisture level for oil expression is expected to exist. In case of rapeseed it is
a moisture level close to 7% (Bargale and Singh, 2000, 130-134). For flaxseed the
optimal moisture content is expected to be around 6% (Zheng and others, 2005, 193-
202). The optimal value for Jatropha is expected to be in the same order of magnitude

Cooking
Cooking causes increased cell wall rupturing thereby facilitating the outflow of oil. The oil
point pressure decreases while pressure build-up increases due to increased viscosity in
turn drastically increasing oil recovery.
3.3.2 Dependent variables
Variables treated in this paragraph change as a result of the independent variables
treated previously.

Oil recovery
This is the most important dependent variable considering the goal of this report. The
effect of the independent variables on oil recovery can be explained by looking at the
dependent variables described below.

Temperature
The friction inside the barrel generates heat which is passed on to the oil. Above certain
levels the amount of phosphor and changes in acid value have a negative effect on
engine performance.

Chapter 3: The screw press


17
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Pressure
When speaking of pressure radial pressure is meant in this report. Pressure is expected
to change due to changes in all independent variables. Pressure is expected to be a
proper measure for oil recovery. In general higher pressure will mean more oil is
recovered.

Throughput
The amount of material transported through the press per unit of time is affected by the
independent variables. Throughput can be used to explain oil recovery and is an
important variable for commercial processing. Low throughputs often imply high oil
recovery.

Energy requirement
Energy requirement is probably related to the pressure inside the press head. The
viscosity of the material to be pressed and the rotational speed will also determine the
amount of energy needed.

Oil point pressure
The oil point is defined as the stage which occurs just before oil flows out of compressed
solid particles (Sukumaran and Singh, 1989, 77-84). The oil-point is theoretically coupled
to the kernel density of an oilseed (Faborode and Favier, 1996, 335-345). As this stage
will always occur before oil is separated from its surrounding solids Sukumaran & Singh
(1989) identified it as an important material property when studying mechanical oil
expression. The material to be pressed consists of oil bearing solids. For the oil to be
freed from the solids it has to come to the surface of the seeds and fill interparticle voids
(Sukumaran and Singh, 1989, 77-84). Oil recovery is assumed to depend on the
pressure exerted on the material in excess of the pressure that is required to reach the
oil point. Sukuraman & Singh found that the cell walls were still in tact at the oil-point
confirming that rupturing of cell walls is not a precondition for oil expression. The oil can
leave the cell through plasmodesmata (pores) in the cell wall (Faborode and Favier,
1996, 335-345). For rapeseed the oil point pressure was found to be 57-97 bar
depending on moisture content and rate of deformation.

The oil point pressure is affected by the rate of deformation in mm/min (which can be
translated to screw speed in case of an expeller) and the moisture content. Press tests by
Sukuraman & Singh (1989) using a loading piston indicate that pressure corresponding to
the oil point increases with increased moisture content and with increased rate of
deformation. The most significant effect can be attributed to moisture. Should the same
principles hold for an expeller it is to be expected that the oil point pressure of the
material inside will increase for higher moisture level and screw speed. Size reduction
and heating can be applied to increase compressibility of the seed material and thus
lower the oil-point pressure (Faborode and Favier, 1996, 335-345).
Chapter 4: Materials and method


18
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
4 Materials and method
This chapter describes all experimental methods conducted at both Eindhoven University
of Technology and Diligent Tanzania LTD. In paragraph 4.1 the experiments conducted at
the university are described. Paragraph 4.2 elaborates on tests done at Diligent Tanzania
LTD.
4.1 Experiments at Eindhoven University of Technology
4.1.1 Materials
Seeds of Jatropha Curcas L were obtained from Diligent Energy Systems (Eindhoven, The
Netherlands). The origin of the seeds is the northern region of Tanzania around Arusha.
Two different batches were supplied; one harvested in July 2005 and the other one in
July 2004. The oil content of the seeds, determined by soxhlet extraction with petroleum
ether, was found to be 37.8 wt.% (dry basis). Further details on the soxhlet extraction
are given in paragraph 4.1.4. Standard Silica crystals were available at the faculty of
Chemical Engineering. Petroleum ether (boiling range 40-60˚C) was bought from (Fluka
Riedel de Haën, Buchs, Switzerland).
4.1.2 Experimental setup
A Danish screw press (BT Bio Presse Type 50, BT biopresser aps, Dybvad, Denmark) was
made available by Diligent Energy Systems and FACT Foundation (Eindhoven, The
Netherlands) to extrude the oil from the seeds in one pass. Figure 4-1 shows the BT 50
test setup. This small capacity press was designed for smaller seeds like rapeseed and
sunflower but proved to run properly on the bigger Jatropha seeds. The capacity of the
press for Jatropha is approximately 12kg/h. The screw is powered by a 1.1 kW electric
motor. Screw RPM is directly regulated between 0-70 RPM by a variable speed
drive(Altivar 31, Schneider Electric SA, Rueil-Malmaison, France). The variable speed
drive is also used for readings on power consumption of the electric motor. The
restriction size at the press cake outlet can be varied by placing different sized nozzles.
Detailed specifications of the press geometry are tabulated in Appendix C. During press
tests only a single screw geometry, which was not optimized for Jatropha, was used.
Pressures during testing reached up to 300 bars at temperatures of over 140 ˚C.




Figure 4-1 Left: BT50 test setup at Eindhoven University of Technology. Middle: situation of
thermocouples and pressure sensor (blue wire). Right: front view inside the press head.

The seeds underwent different pre-treatments in order to evaluate the effect on amongst
others oil yields and oil quality. Hull content was varied between 60 and 100% using the
custom built de-huller shown in Figure 4-2. Separation of hull and kernel had to occur
manually. The prepared samples each weighing about 2 kg were either sealed for storage
or instantly fed into the screw press. Determination of the moisture content is done by
oven drying (Heraeus BR 6000, Hanau, Germany) just before the seeds are fed to the
screw press in order to minimize errors.
Chapter 4: Materials and method


19
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

Figure 4-2 Hand powered de-huller designed
together with Gerard van Hout from the
workshop at Mechanical engineering.

4.1.3 Experimental procedure
The description of a typical experiment is used to explain the experimental procedure. To
prevent jamming the press head is heated to 70-80˚C using a hot air pistol. While
continuing heating small amounts of seeds are fed into the hopper. A sample size of 2kg
was chosen as this allows stable measurements and simplifies the calculation of
throughput. Crude oil is collected at the oil outlet and press cake exit on the left (see
Figure 4.1 middle) in the shape of small pellets. Temperature is measured at 4 positions
along the press cage (type J Thermal Insulated thermocouple, Themocoax, Netherlands).
The first is located at the press cake exit, the second at the point where maximum
temperature is expected and the third & fourth at different positions along the oil outlet
as indicated in Figure 4-3. The radial pressure is measured (6152AAA Mold Cavity
Pressure Sensor, Kistler Instruments AG, Switzerland) at the point of maximum expected
pressure, which coincides with the point of maximum temperature. The sensor signal is
amplified into a signal in the range of 0-10V by a charge amplifier (5073 Industrial
Charge Amplifier Manufacturing, Kistler Instruments AG, Switzerland). All data was
digitally recorded using a PCMCIA card (NI4350, National Instruments, United States)
and ‘Virtual Bench Logger’. In addition the time required for a batch to pass through the
press was measured. It was difficult to determine the instant at which the pressing
finished, therefore the time measurement was started when the pressure showed a steep
increase after seeds were fed and stopped once pressure drastically declined while the
press was almost empty. Power consumption was recorded using one of the Altivar 31
voltage output signals. The output signal ranged from 0-10V where 10V indicates twice
the nominal power of the 1.1kW electric motor (Altivar 31 user manual). A graph in
appendix D shows maximum power measurement while the press is jammed.

Jatropha seeds with increased different moisture content were obtained by equilibrating
the seeds for one week in a sealed environment with a fixed amount of water inside
plastic bag. A similar method was used in other studies on mechanical expression of oil
seeds (Vadke and Sosulski, 1988, 1169-1176;Zheng and others, 2005, 193-202). To
decrease the moisture content below the natural value of around 7% a similar method
was used with silica crystals. The silica was oven dried overnight at 103˚C. All moisture
contents were determined according to DGF standard method B-I 4 (DGF-
Einheitsmethoden, 2002).
Chapter 4: Materials and method


20
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Point 4: press
cake exit
Point 1
Point 2: oil outlet
Point 3: maximum
temperature and pressure

Figure 4-3 Location of measurement devices

After pressing a sample of the press cake was analyzed on oil content with Soxhlet
extraction according to DGF standards (DGF-Einheitsmethoden, 2002). After leaving the
oil to settle for a week relative amount of foots in the crude oil could be determined.
4.1.4 Analysis
Oil content
Prior to determining the oil content of seed or press cake the sample was dried at 103˚C
for 5 hours in a drying oven (Heraeus BR 6000, Hanau, Germany). The oil contents are
compared on a dry basis for ease of comparison. After drying the sample was ground
together with a small amount of petroleum ether for 20 seconds at 25Hz in a ball mill
(MM301, Retsch GmbH & Co, Haan, Germany). Oil content was determined through
overnight Soxhlet extraction with petroleum ether with the setup in Figure 4-4. The
petroleum ether was heated to 100°C to ensure sufficient solvent flow rate. Standard
cellulose extraction thimbles (603 Cellulose Extraktionshülsen, Schleicher & Schuell
Microscience GmbH; Whatman group, Dassel, Germany) were used to prevent solid
particles from entering the oil sample. After extraction the obtained oil sample was oven
dried at (103 ± 2) °C for 5 hours to remove remaining solvent. The oil content,
expressed in g/100g, was determined by comparing the weight of the extracted oil to the
total weight of the sample prior to extraction.


Figure 4-4 Soxhlet extraction setup at Eindhoven University of Technology

Chapter 4: Materials and method


21
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Moisture content
A drying method according to DGF standards was used to determine the moisture content
of the seeds before pressing. The drying method is similar to the one used before soxhlet
extraction. Samples of approximately 5 grams were weighed and put in the oven
(Heraeus BR 6000, Hanau, Germany) for 5 hours. After drying the weight was again
determined.

Oil analysis
Analyses of some oil samples on phosphorous content and acid value was out contracted
to consulting engineers (Ingenia, Eindhoven, The Netherlands). Analysis of the samples
was done by ASG Analytik-Service Gesellschaft mbH (Neusäss, Germany) according to
DIN methods. The samples consisted of crude Jatropha oil that had been subjected to
different cleaning methods.
4.1.5 Accuracy and reproducibility of experimental results
Results obtained at Eindhoven University of Technology are more accurate than the ones
from Tanzania, due to superior equipment. Press experiments were only conducted once
due to time constraints. Repetition of measurements would have increased reliability and
accuracy. Soxhlet extraction can be repeated at 1% accuracy according to DGF
standards. More information on accuracy is written in paragraph 5.1.
4.2 Experiments at Diligent Tanzania LTD
As official descriptions of some tools were not available pictures and additional
information have been included in Appendix E.
4.2.1 Materials
Seeds from Jatropha Curcas L were obtained from Diligent Tanzania (Arusha, Tanzania).
All the seeds used for this research originated from the same village Engaruka located
180km North West of Arusha in order to create homogeneous samples. The oil content of
the seeds, determined by soxhlet extraction with petroleum ether and hexane, were
found to be 38 wt.% and 39 wt.% successively (dry basis). The moisture content of
stored seeds during the testing period was determined to be 6.7 wt.%.
4.2.2 Experimental setup
Oil expelling tests were performed using a Sayari expeller (Vyahumu Trust, Morogoro,
Tanzania). Although the original design of this strainer press, known under the name
Sundhara, originates from India it was adapted to Tanzanian standards by engineers
from GTZ in the early 90’s. The Sayari press, depicted in Figure 4-5, is now a locally
produced screw press that was initially designed to process sunflower seeds although
suitability to press Jatropha is mentioned in the ‘Operator’s instruction manual’.
Production is subcontracted to three workshops located close to the Vyahumu office, one
of which is Intermech (More information on the workshop and Sayari expeller, including
pictures showing the assembly of the Vyahumu press is shown in Appendix F). The
capacity of the Sayari for sunflower is around 140kg/hr (Peter Chisawillo, Director
Intermech) and the minimum processing capacity is 70kg/hr (Operator’s instruction
manual). According to Vyahumu and Intermech (Lehada C.Shila, 2007;Peter Chisawillo,
2007) the capacity for Jatropha should be about 70-80 kg/hr. Tests at Diligent Tanzania
LTD showed similar results. The press is powered by a 5.5 kW asynchronous generator
(MA 132 SA 4, Marelli Motori, Italy) rotating at 1500 RPM. A simple transmission made of
sprockets and chains reduces the rotational speed of the press worm to 55 RPM. The four
cones mounted on the screw (see Figure 4-5 right side) enable pressure to build up by
gradually reducing the clearance between screw and press cage. Pressures inside the
press can reach values up to 150 bars at temperatures of around 110˚C. High pressure
and friction inside the press cause significant heating of the press during operation.

Chapter 4: Materials and method


22
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries


Figure 4-5. Left: Sayari press at Diligent Tanzania premises, with (1) feeder, (2) oil outlet, (3)
press cake outlet. Right: Worm inside Sayari screw press showing 3 cones with diameters 90,
94 and 98mm (4) from right to left. The fourth cone, located at the press cake outlet, is
adjustable by a hand operated turning wheel on the far left.

4.2.3 Experimental procedure
The description of a typical experiment is used to explain the experimental procedure.
Before starting the measurement the press required preheating to 60-70˚C during
approximately 20 minutes of normal operation using seeds, press cake or a mixture of
the two. After preheating seeds or press cake were inserted in the feeder (1) on the right
side of the press shown in Figure 4-5. The normal seed batch size for measurement was
10-20kg because it takes some time for the process to stabilize. Crude oil was collected
at the oil outlet (2) and press cake exits the machine at the adjustable cone (3). A
thermocouple (type J Thermal Insulated thermocouple, Themocoax, Netherlands) located
at the third cone (see Figure 4-5 (4)) measured temperature changes. The radial
pressure was measured at the same cone (6152AAA Mold Cavity Pressure Sensor, Kistler
Instruments AG, Switzerland). A charge amplifier (5073 Industrial Charge Amplifier
Manufacturing, Kistler Instruments AG, Switzerland) transformed the sensor signal to a
signal in the range of 0-10V. All data was digitally recorded using a PCMCIA card
(NI4350, National Instruments, United States) and ‘Virtual Bench Logger’. In addition the
time required for a batch to pass through the press was measured. It was difficult to
determine the instant at which the pressing finished, therefore the time measurement
was started when the pressure showed a steep increase after seeds were fed and
stopped once pressure drastically declined while the press was almost empty in order to
generate comparable results.

After pressing a sample of the press cake was analyzed on oil content with Soxhlet
extraction according to DGF standards (DGF-Einheitsmethoden, 2002). After leaving the
oil to settle for a week relative amount of foots in the crude oil could be determined. The
remaining press cake was fed back into the press and the whole process repeated for a
second time.

Jatropha seeds with increased moisture content were obtained by equilibrating the seeds
for one week in a sealed environment with a fixed amount of water inside a plastic
container. Other studies on mechanical expression of oil seeds (Vadke and Sosuslski,
1988; Zheng, Wiesenborn et al 2004) reported to have used similar methods. Moisture
content was reduced below the natural 6.7% by sun drying. As this method depends
Chapter 4: Materials and method


23
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
heavily on weather conditions the drying time was not constant. Resulting moisture
contents were determined by oven drying.
4.2.4 Analysis
Oil content
Oil content was again determined through soxhlet extraction with hexane. The oil
content, expressed in g/100g, was determined by comparing the weight of the extracted
oil to the total weight of the sample prior to extraction. The method differed from DGF
standards as some of the required equipment was not available in Arusha. The setup
differed from DGF standard on the following aspects:

• A mortar and pestle were used to grind samples prior to extraction instead of the
ball mill.
• Samples were wrapped in normal filter paper (Whatman qualitative filter paper
125mm, Cat No. 1001125) instead of extraction thimbles.
• Hexane was used as solvent instead of petroleum ether. However comparative
tests for Jatropha with petroleum ether showed comparable results with hexane
the values were 38% and 39% respectively.
• The electrical balance (440-33N, Gottl. Kern & Sohn GmbH, Balingen-Frommern,
Germany) did not meet accuracy requirements; 0.01g instead of 0.001g.
• Samples of ± 10 grams were used instead of the prescribed ± 5 grams in order to
reduce the effect of measurement errors.
• The samples were not dried before extraction. A correction was applied afterwards
based on drying of representative samples.

These deviations from the standard method lessen the accuracy of the results. Due to the
less accurate soxhlet setup in Tanzania samples were transported to the Netherlands for
accurate extraction.

Moisture content
A drying method according to DGF standards was used to determine the moisture content
of the seeds before pressing. Samples of approximately 5 grams were weighed and put in
the oven (Nevica, designed for domestic use, see Appendix E for picture and drying
profile) for 3 hours. Due to the simple oven design a drying profile had to be determined
in order to keep the temperature in the acceptable range of 80-100˚C.

Foots
The relative amount of foots in the crude oil were determined after one week of settling
in plastic bottles (see Figure 4-6). Oil and foots levels were marked and refilling with
known quantities of water resulted in an approximation of the actual volumes. A 1000ml
graduated measuring cylinder with 5ml accuracy was used for the water refilling.


Figure 4-6 Oil samples left for settlement
Chapter 4: Materials and method


24
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
4.2.5 Accuracy and reproducibility of experimental results
Repetition of each pressing experiment would allow proper judgment of accuracy. Due to
limitations in both means and time repetition was seldom possible for the press tests
executed in Tanzania. Results on oil yield, temperature are therefore meant to identify
trends and are assumed to deviate up to ±2 and ±5%. For example oil content found to
be 37% means that the actual value lies between 35-39%. Less accurate mounting of the
pressure sensor in the Sayari expeller compared to the BT50 caused pressure values to
deviate ±20% instead of 10% as will become clear from the results section 5.1. In
addition to the less accurate mounting of the pressure sensor its position relative to the
third cone changes up to 1cm when changing restriction size. To what extent this affects
pressure measurements could not be determined. The accuracy for measuring oil and
foots levels was estimated at ±50ml, due to inaccurate reading of the marked bottles.
This comes down to deviation in the range of 5-15% depending on sample size.
Chapter 5: Results and discussion


25
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
5 Results and discussion
The objective of this chapter is to show the effect of various seed pre-treatments and
press settings on the oil yield of mechanical expression of Jatropha seeds. Increased oil
yield can be achieved either by increasing the amount of crude oil or by reducing the
relative amount of foots in the oil. The variables included in this research are listed below
in Table 5-1. An overview and explanation of the exact values for speed, moisture
content etc is given in Appendix G.

Table 5-1 Overview of the tested variables for the two different presses
BT50 Sayari
Press settings
RPM X
Restriction X X
Seed pretreatments
Moisture content X X
Dehulling X X
Torrefaction X
Heating in water X X
Crushing X X



The effect of a pre-treatment on the pressing process is best explained by the behaviour
of pressure and temperature inside the press (Eggers, Broeck, and Stein, 2006, 494-
499;Vadke and Sosulski, 1988, 1169-1176;Willems, 2007). The following paragraphs
provide an overview of the most relevant test results and an explanation of what causes
the observed differences. Finally the experimental results are used to optimize the
pressing process.

The test settings for the two different press types that were used are as follows. For the
BT 50 all tests with respect to seed pre-treatment were conducted at standard conditions
which are 49 RPM, 100% hull and a restriction diameter 9mm. The standard settings for
the Sayari expeller were 55 RPM, 100% hull and a restriction gap size of 1.8mm. These
settings were selected as standard setting because they ensured smooth running of the
presses. The restriction sizes of the two presses schematically indicated in
Figure 5-1 refer to different geometries and can therefore not directly be compared.

1.8 mm 9 mm


Figure 5-1 Schematic representation of the restriction shape of the Sayari expeller (left) and the
BT50 (right). The white circle on the left side represents the adjustable cone of the Sayari expeller.

All tests in the BT50 press were single pass, meaning that only seeds were pressed and
the press cake was not fed back for a second run. Most tests in the Sayari expeller were
dual pass tests. Because of the higher residual oil content of the press cake from the
Sayari expeller a second press run was worthwhile. Single pass and dual pass are
indicated in the figures with Sayari results.
Chapter 5: Results and discussion


26
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
5.1 Data processing
This paragraph briefly explains the coming about of the graphical representations in the
following paragraph. The main points concerned here are data selection and error margin
determination

Oil recovery
The oil recovery graphs were obtained by comparing the residual oil content in the press
cake to the initial oil content in the seeds using the following formula (see Appendix H for
derivation):

(
(
(
(
¸
(

¸

|
¹
|

\
|

|
¹
|

\
|

− =
Wo
Wo
Ws
Ws
Y
1
1
1
with Y= oil yield or oil recovery (%oil/oil)
Wo= Oil content in original seed (g/100g)
Ws= Oil content in press cake sample (g/100g)

The oil recovery is expressed as the recovered fraction of the initial amount of oil. So
knowing that the Jatropha seed contains 37wt.% oil and assuming an imaginary recovery
rate is 80% this means that from 100kg seeds 29.6kg of oil are recovered. Taking into
account the density of 0.92kg/l (Akintayo, 2004, 307-310;Pramanik, 2003, 239-
248;Ruud van Eck, 2006) this means 32.2 litre.

The error bars indicate a ± 2% fixed error, which means that a value of for example 37%
is between 35-39%. The accuracy of the soxhlet extraction is stated as 1% (DGF-
Einheitsmethoden, 2002). The other 1% is attributed to variations in the oil content in
the press cake when it leaves the machine.

Throughput
Throughput was determined by dividing the batch weight by the time required for
pressing the batch. Starting time was identified by a sudden increase in pressure. Time
recording is slightly inaccurate as it is hard to determine at what moment the last
material exits the press. Timing errors are assumed to be within 30 seconds leading to
5% inaccuracy for large batches (BT50 tests with 2kg batches and Sayari tests with 20kg
batches) and 10% inaccuracy for small batches (BT50 tests with 1kg batches and Sayari
tests with 10kg batches).

Pressure
The graphs indicating pressure dependence were compiled using average pressure
values. Only pressure values after a sudden rise, indicating pressure build up inside the
press, were included. Due to the rotating screw the pressure occurring in the press head
is very turbulent. In spite of the Kistler 6152AAA pressure sensor being 1% accurate a
10% error is used to take the fluctuations into account. Graphs showing the initial
measurement output are shown in Appendix I to support this choice.

Temperature
The temperatures as indicated in figures similar to Figure 5-2 b represent the average
value at the same axial location as the pressures. Thermocouple calibration in a
thermostatic bath showed an accuracy of 2.5% in the temperature range of 90-100°C
and less than 1% for lower temperature (see Appendix D). The available recording
equipment lead to the selection of thermocouples that were suboptimal for the
temperature range and therefore the variation in temperature is assumed to be 5%.


Chapter 5: Results and discussion


27
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Energy input
The graphs for energy input were computed from averaged power output readings of the
Altivar 31 variable speed drive converted into kJ/kg and kJ/l values. As the accuracy was
not given by the producer a 5% error was assumed for both graphs. Unfortunately the
energy input could not be measured during the experiments carried out on the Sayari
expeller in Tanzania. An approximation of the energy requirement for the BT50 made
prior to testing is shown in Appendix J.
5.2 Influence of RPM

BT50
The graph in Figure 5-2 a shows a reduction in oil recovery for increased rotational
speed. This effect is largely explained by the increasing throughput indicated in the same
figure. Increased throughput means reduced residence time and thus less chance for the
oil to flow from between the solid material. The higher residual oil content in the material
ensures that the viscosity of the paste remains relatively low and therefore pressure
build-up is also lower at higher speed as can be seen in Figure 5-2 b. The lower pressure
is a second cause for the oil recovery to be lower for higher speeds. A third cause for
decrease in oil recovery with increasing screw speed can also be an increase in oil-point
pressure as studies by Sukurman & Singh (1981). The rotational speed of the worm
inside the screw press could be considered analogue to the rate of deformation under
uni-axial compression. Under uni-axial compression the oil-point pressure increases with
increasing rate of deformation (Sukumaran and Singh, 1989, 77-84). Only the pressure
exerted in excess of the oil-point pressure is assumed to result in oil to be recovered
from the compressed seeds. Applying the same assumption to screw pressing would
predict lower oil yield at higher screw speed.

Combining throughput and oil recovery reveals that the oil production capacity varies
from 4.2l/hr at 28RPM to 8.7l/hr at 70RPM. The increase in temperature for higher speed
could be explained by increased friction heating or simply be due to the fact that the
tests were executed successively from low to high speed and the press had not yet
reached equilibrium temperature. Temperature differences can be considered insignificant
for changes in rotational speed.

20 30 40 50 60 70
0
20
40
60
80
100
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Screw speed (RPM)
Oil recovery
0
2
4
6
8
10
12
14
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)
Throughput

20 30 40 50 60 70
0
20
40
60
80
100
120
140
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Screw speed (RPM)
Pressure
0
20
40
60
80
100
120
140
Temperature

Figure 5-2 a Influence of screw speed on oil
recovery and throughput for the BT50
Figure 5-2 b Influence of screw speed on radial
pressure and temperature for the BT50

The energy requirement to press the seeds is represented in Figure 5-2 c. Higher
rotational speed requires more power and at the same time reduces processing time. The
negative slope of the upper line indicating energy input per litre in Figure 5-2 c shows
Chapter 5: Results and discussion


28
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
that the energy requirement is most strongly affected by the reduction in processing
time.

20 30 40 50 60 70
100
150
200
250
300
350
400
E
n
e
r
g
y

i
n
p
u
t

(
k
J
/
l
)
E
n
e
r
g
y

i
n
p
u
t

(
k
J
/
k
g
)
Screw speed (RPM)
Energy input per kg seed
100
150
200
250
300
350
400
Eenrgy input per liter crude oil

Figure 5-2 c Influence of screw speed on energy requirement
for the BT50 expressed per kg and per litre crude oil

5.3 Influence of restriction size

BT50
Figure 5-3 a and Figure 5-3 b show the influence of the choke opening on pressure and
residual oil content for the BT50 press. Decreasing the choke opening increases the
resistance to flow at the end of the extruder. The pressure required to overcome this
resistance significantly increases judging from Figure 5-3 b. It is difficult to draw
conclusions on the effect on throughput as only two data points are available. Judging
from the figures the higher oil recovery for smaller restriction can directly be attributed to
increased pressure and friction.

6 8 10 12
0
20
40
60
80
100
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Restriction size (mm)
Oil recovery
0
2
4
6
8
10
12
14
16
Throughput
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)

6 8 10 12
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Restriction diameter (mm)
Pressure
0
20
40
60
80
100
120
140
Temperature

Figure 5-3 a Influence of restriction size on oil
recovery and throughput for BT50 press

Figure 5-3 b Influence of restriction size on
temperature and pressure for BT50 press


Because the throughput for 7mm was not measured it is also not possible to calculate the
required energy input. The two data points in Figure 5-3 a incline to a slight decrease of
throughput with reduced restriction size due to the balancing effect of increased
Chapter 5: Results and discussion


29
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
resistance and better filling of the press chamber. Although limited data is available on
the effect of restriction diameter on energy requirement for the BT50 Figure 5-3 c
suggests a reversed dependence between restriction size and energy input.

8 10 12
0
50
100
150
200
250
300
350
400
I
n
p
u
t

e
n
e
r
g
y

(
k
J
/
l
)
I
n
p
u
t

e
n
e
r
g
y

(
k
J
/
k
g
)
Restriction diameter (mm)
Energy input per kg seed
0
50
100
150
200
250
300
350
400
Energy input per liter crude oil

Figure 5-3 c Influence of restriction size on energy requirement
for the BT50 expressed per kg and per litre crude oil

Sayari
As can be seen in Figure 5-4 a reducing restriction size clearly results in an increase in
the amount of oil recovered from the seeds. The effect of restriction size was only
observed for a single pass as this already shows how this variable affects oil yield. For a
second pass the shape of the oil recovery graph is expected to be similar in shape
although less steep. The expected reduction in steepness is due to being closer to the
theoretic maximum oil recovery at smaller restriction sizes after the first pass. Obviously
when decreasing the restriction size the mixture inside the press experiences more
resistance when exiting the press cake outlet.

0 1 2 3 4 5
0
20
40
60
80
100
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Restriction (mm)
Oil recovery single pass
40
50
60
70
80
90
100
110
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)
Throughput single pass

0 1 2 3 4 5
0
20
40
60
80
100
120
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Restriction (mm)
Pressure single pass
0
10
20
30
40
50
60
70
80
Temperature single pass

Figure 5-4 a Influence of restriction size on oil
recovery and throughput for Sayari press
Figure 5-4 b Influence of restriction size on temperature
and pressure for Sayari press


The supply of input material is assumed to stay more or less constant resulting in
increased compression close to the press cake outlet as is indicated by the pressure
development in Figure 5-4 b. The increase in throughput with decreasing restriction size
Chapter 5: Results and discussion


30
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
can appear conflicting at first sight. It might be contributed to better filling of the voids
inside the worm channel at higher pressure therefore enhancing the press’s performance.
The increase in temperature with decreasing restriction size is easily explained by
increased resistance and pressure. Although oil recovery is highest at smaller restriction
size there is an operational limit. The tightest setting of 0.6mm exceeded the operational
limit of the Sayari press resulting in severe jamming. For this reason corresponding
values for throughput, temperature and pressure are not indicated in Figure 5-4 a and
Figure 5-4 b.
5.4 Influence of moisture content
BT50
The results of press tests for varying moisture content using the BT50 press are shown in
Figure 5-5 a to Figure 5-5 c. Figure 5-5 a shows that with increasing moisture content
the throughput decreases, although the decrease is less than with increasing screw
speed. The effect of moisture content on oil recovery is on the other hand considerable.
Only part of the decrease in oil recovery with increased moisture content can be
attributed to the increase in throughput. When comparing Figure 5-5 a to Figure 5-2 a it
is apparent that with increasing moisture content the increase in throughput is smaller
and the decrease in oil recovery is bigger, which directly points out that changes in
throughput caused by varying moisture content are an insufficient explanation for the
occurring changes in oil recovery. The strong decrease in pressure visible in Figure 5-5 a
hints another cause for the reduction in oil recovery. Higher moisture content might
cause increased deformability and reduced rupture fore as suggested by Singh & Bargale
(2000). In addition emulsification or plasticizing effects occurring in the pressed material
at higher moisture levels reduce the viscosity and thereby decrease pressure build-up.
This pressure drop seems an appropriate explanation for the rapid decline in oil recovery
in Figure 5-5 a. The suggestion by Willems (Willems, 2007) that during expression water
can only be removed by evaporation brings up another contribution to the cooling effect.
The amount of water present in the seeds however seems insufficient to induce the
temperature drop by more then 50%. A more plausible explanation is a cooling effect of
the input material. Lower friction and pressure build-up due to the reduced viscosity
reduce heat generation at higher moisture content. The cooling effect of the cold material
entering the press might in this case counterbalance heat generation or even lead to
faster cooling than is the case after normal experiments. Although oil recovery is optimal
for a moisture content of 2.14% this might not be the optimal processing conditions as
the oil temperature reaches high values that could lead to changes in the composition of
the oil as is the case for rapeseed (Singh and Bargale, 2000, 75-82;Zheng and others,
2005, 193-202).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
0
20
40
60
80
100
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Moisture content (wt% w.b.)
Oil recovery
0
2
4
6
8
10
12
14
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)
Throughput

2 4 6 8 10 12 14
0
50
100
150
200
250
300
350
P
r
e
s
s
u
r
e

(
b
a
r
)
Moisture content (wt% w.b.)
Pressure
0
20
40
60
80
100
120
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
Temperature

Figure 5-5 a Influence of moisture content on oil
recovery and throughput for the BT50 press
Figure 5-5 b Influence of moisture content on
temperature and pressure for the BT50 press
Chapter 5: Results and discussion


31
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

Judging from Figure 5-5 c the energy requirement per kilogram of input material reduces
for increasing moisture content. However at higher moisture content more seeds are
required judging from Figure 5-5 a. More seeds per litre mean longer processing time per
litre. This results in the decreased slope of the upper line in Figure 5-5 c at higher
moisture levels. From a cost perspective it is not desirable to decrease the energy use by
increasing the moisture level in the seeds as the costs of three times as much input
material per litre are much higher than the savings of using three times less energy. This
example directly shows that energy requirement is not expected to be a limiting factor.

2 4 6 8 10 12 14
0
100
200
300
400
500
600
700
800
900
I
n
p
u
t

e
n
e
r
g
y

(
k
J
/
k
g
)
Moisture content (wt%)
Energy input per kg seed
0
100
200
300
400
500
600
700
800
900
Energy input per liter crude oil
I
n
p
u
t

e
n
e
r
g
y

(
k
J
/
l
)

Figure 5-5 c Influence of moisture content on energy requirement
for the BT50 expressed per kg and per litre crude oil

Sayari
The effect of moisture content was tested for dual passing in the Sayari expeller. Due to
the design of the Sayari expeller it is much more difficult to execute steady test runs. In
addition conditioning of the seed moisture level was less accurate due to less
sophisticated instruments. The uneven distribution of data points in Figure 5-6 a is the
result of this unsteady operation and imprecise conditioning. Oil recovery after dual
passing shows a slight decrease for increasing moisture content. The effect of moisture is
less than for the BT50. An explanation could lie in the multiple stage compression in the
Sayari expeller. Although data on throughput seem inaccurate the increase in throughput
with moisture level is plausible when compared tot the BT50 results in Figure 5-5 a.
Temperatures and pressures are also less aligned compared to the previous figures. Test
results from the Sayari expeller in Figure 5-6 b show a drastic decrease in pressure with
moisture content, as was the case for the BT50, and to a lesser extend temperature.
Judging from Figure 5-5 b and Figure 5-6 b for moisture levels between 6 and 9% the
temperatures in the Sayari press are about 15% lower than for the BT50. The slopes of
both temperature lines are comparable. In the same moisture range the pressure
measured in the BT50 is approximately two times the value measured in the Sayari
expeller. This shows that changes in pressure are an indicator for changes in
temperature, but the absolute temperature value is largely determined by the combined
effect of pressure, friction, material and design of cage and screw.

Chapter 5: Results and discussion


32
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
6.5 7.0 7.5 8.0 8.5 9.0 9.5
0
20
40
60
80
100
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Moisture content (wt% w.b.)
Oil recovery dual pass
Oil recovery single pass
40
50
60
70
80
90
100
110
120
130
140
150
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)
Throughput of seeds
average single and dual pass

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
0
10
20
30
40
50
60
70
80
90
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Moisture content (wt% w.b.)
Pressure single pass
0
20
40
60
80
Temperature single pass

Figure 5-6 a Influence of moisture content on oil
recovery and throughput for the Sayari press

Figure 5-6 b Influence of moisture content on
temperature and pressure for the Sayari press
(single pass)
When comparing Figure 5-6 b and Figure 5-6 c one can see the difference between
pressing seeds (single pass) and pressing press cake (dual pass). What stands out is that
the temperature at dual pass is higher while the pressure is lower. The higher
temperature is the result of repressing the press cake while it is still hot (about 60-80°C).
Due to this unintended preheating of the press cake comparison between pressing seeds
and press cake is difficult.
6.5 7.0 7.5 8.0 8.5 9.0 9.5
0
10
20
30
40
50
60
70
80
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
d

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Moisture content (wt% w.b.)
Pressure dual pass
0
20
40
60
80
100
Temperature dual pass

Figure 5-6 c Influence of moisture content on
temperature and pressure for Sayari press (dual pass)
5.5 Influence of hull fraction

BT50
As the oil-point pressure is lower for softer oilseeds the same will probably hold for de-
hulled seeds compared to normal seeds as the bulk material becomes softer (Faborode
and Favier, 1996, 335-345). Based on this assumption a higher oil recovery is expected
although Figure 5-7 a clearly shows the opposite. What happens when de-hulling the
seeds is that the material inside the press becomes fattier and due to a difference in
shape friction between the mixture and the barrel drops below the friction between the
mixture and the worm. When the hull content is reduced to 66% of its original value
causes slip in the feed section and the solid-oil mixture starts to stick to the worm. The
result is a sharp decrease in throughput (Figure 5-7 a) which can even drop to zero. The
Chapter 5: Results and discussion


33
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
same problem was also brought forward by Ward (1976) when dealing with high oil
content seeds. Khan & Hanna (1983) suggest the use of a low friction shaft combined
with higher friction cage bars for softer seeds like de-hulled Jatropha. Decrease in
pressure and temperature are as before due to reduced friction (Figure 5-7 b).

65 70 75 80 85 90 95 100 105
0
20
40
60
80
100
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Hull content (wt%)
Oil recovery
0
2
4
6
8
10
12
14
Throughput
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)

60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
P
r
e
s
s
u
r
e

(
b
a
r
)
Hull content (wt%)
Pressure
0
10
20
30
40
50
60
70
80
90
100
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
temperature

Figure 5-7 a Influence of hull content on oil
recovery and throughput for BT50 press
Figure 5-7 b Influence of hull content on
temperature and pressure for the BT50 press

The required energy input decreases for lower hull content as indicated by Figure 5-7 c.
The data points for 66% hull content are too high due to slip and rotation movement of
the solid-oil mixture instead of horizontal conveyance. Throughput reduced to zero during
the experiment thereby causing the deviant result. Difficulties experienced during testing
demonstrate that in standard screw presses a reduction of the hull fraction is applicable
up to 80% of the original hull fraction. The combined effect of a decrease in power
consumption (expressed in Watt) and an increase in throughput at comparable oil yield is
increased energy efficiency of the process (less kJ/litre) when hull fraction is reduced to
80%. Below a certain threshold value between 66-80% hull the press performance is
sharply reduced due to deteriorating mixture conveyance.

65 70 75 80 85 90 95 100 105
0
100
200
300
400
500
600
700
800
900
1000
1100
E
n
e
r
g
y

i
n
p
u
t

(
k
J
/
l
)
E
n
e
r
g
y

i
n
p
u
t

(
k
J
/
k
g
)
Hull content (wt%)
Energy input per kg seed
0
200
400
600
800
1000
Energy input per liter liter crude oil

Figure 5-7 c Influence of hull content on energy requirement
for the BT50 expressed per kg and per litre crude oil

Sayari
The test results on hull content for the Sayari expeller were only limited. Nevertheless
the results in Figure 5-8 a and Figure 5-8 b indicate a dependence on hull fraction
Chapter 5: Results and discussion


34
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
comparable to the BT50 press. Only dual pass values are given as they tend to show less
variance than single pass values as was also observed in Figure 5-6 a.
60 70 80 90 100
0
20
40
60
80
100
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Hull content (wt%)
Oil recovery dual pass
50
60
70
80
90
Throughput average
single and dual pass

60 70 80 90 100
0
10
20
30
40
50
60
70
80
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Hull content (wt%)
Pressure dual pass
0
10
20
30
40
50
60
70
80
90
100
Temperature dual pass

Figure 5-8 a Influence of hull content on oil
recovery and throughput for Sayari press
Figure 5-8 b Influence of hull content on
temperature and pressure for the Sayari press
5.6 Influence of seed preheating

BT50
Heating of the input material can be applied to increase compressibility of the seed
material and thus lower the oil-point pressure (Faborode and Favier, 1996, 335-345).
The effect of preheating of the seeds before pressing was tested by preheating standard
seeds (6.7% moisture) during 30 minutes at various temperatures. Moisture content of
the seeds was not determined before pressing. The inner structure of the Jatropha seeds
was not visibly affected after heating. The influences on oil content and temperatures
inside the press are depicted in Figure 5-9 a and Figure 5-9 b. The effect on oil recovery
and throughput is negligible. Increase in temperature can be explained through a
combination of the heated input material and minor increase in friction as the material is
dryer for higher preheating temperatures.

100 120 140 160 180
0
20
40
60
80
100
T
h
r
o
u
g
h
p
u
t

(
k
g
/
h
r
)
O
i
l

r
e
c
o
v
e
r
y

(
%
o
i
l
/
o
i
l
)
Seed temperature (degrees Celsius)
Oilrecovery
0
2
4
6
8
10
12
14
Throughput

100 120 140 160 180
0
20
40
60
80
100
120
140
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
P
r
e
s
s
u
r
e

(
b
a
r
)
Seed temperature (degrees Celsius)
Pressure
0
20
40
60
80
100
120
140
160
Temperatur

Figure 5-9 a Influence of seed input temperature
on oil recovery and throughput for the BT50
press
Figure 5-9 b Influence of seed input temperature
on temperature and pressure for the BT50 press

Chapter 5: Results and discussion


35
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
The assumption of minor friction increase is supported by Figure 5-9 c, which shows a
small increase in the amount of energy required for pressing seed that were pre-heated
at higher temperatures while processing time is constant. This indicates that the engine
encounters slightly more resistance. The upper line in Figure 5-9 c indicates a small
decrease in energy use per litre oil at a preheating temperature of 140°C. The reduction
in energy use would not balance the energy needed for preheating and is thus not
advisable for Jatropha seeds. Furthermore the variation is within the error margin and
therefore uncertain.
100 120 140 160 180
0
100
200
300
400
500
600
700
800
900
E
n
e
r
g
y

i
n
p
u
t

(
k
J
/
k
g
)
Seed temperature (degrees Celsius)
Power
0
100
200
300
400
500
600
700
800
900
Energy input per liter crude oil
E
n
e
r
g
y

i
n
p
u
t

(
k
J
/
l
)

Figure 5-9 c Influence of seed input temperature on energy
requirement for the BT50 press expressed per kg and per litre crude oil
5.7 Influence of other pre-treatments
In addition to the previously mentioned variables some other test were conducted
although in less detail. The results are presented in Table 5-2. Cracking was expected to
decrease energy use and reduce the oil-point pressure (Faborode and Favier, 1996, 335-
345). The results show that the assumptions do not hold for Jatropha as the energy
reduction is negligible and the oil recovery is lower at pressures comparable to the
standard situation. The BT50 data in Table 5-2 suggest that the oil-point pressure is
lower for the cooked samples as their oil recovery is highest while pressure are
significantly lower than those for dry seeds. The cooking pre-treatment also shows a
positive effect in the Sayari expeller.

Table 5-2 General test results for additional experiments on seed pre-treatment. * Samples were
dried to certain moisture content before pressing. Oil recovery values for Sayari expeller are after
dual pass. The temperatures indicated for the Sayari expeller are the highest temperatures that
occurred, which was during dual pass. Boiling showed a very high oil recovery after single pass in
the Sayari press of 88.7%.
Residual oil Oil recovery Press head temperature [°C] Oil temperature [°C] Pressure [bar] Throughput [kg/hr] Energy [kJ/l]
BT50
normal 11.0% 79.2% 98 85 122 9.7 142
crushed 11.9% 77.2% 92 82 120 9.8 139
water 70°C 6.1% 89.0% 120 100 220 9.2 312
water 80°C 6.3% 88.7% 141 115 198 9.4 279
Sayari
normal 6.9% 87.5% 84 - 39 54 -
crushed 7.4% 86.4% 84 - 36 63 -
water 80°C 5.0% 91.2% 103 - 86 69 -

5.8 The effects on foots
A striking characteristic of the crude Jatropha oil obtained by screw pressing is the very
high amount of foots or sediments in the oil. After one week settling time approximately
30 to 50% of the volume consists of sedimentary deposits. For normal filtering
Chapter 5: Results and discussion


36
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
techniques only 5-10% sediment is acceptable (Ferchau, 2000), which is also the amount
of sediments present in most other plant oils after mechanical extraction. It was not
possible to determine if and how pre-treatments affect the amount of sediments. The
table in Appendix K shows more detailed values of the amount of foot for various
samples.
5.9 Oil quality parameters
Oil quality was judged base don the five parameters shown in Table 5-3. Three very
diverse samples of filtered oil were analyzed by ASG (Analytik-Service Gesellschaft mbH,
Neusäss, Germany). These samples cover both the widest range in oil temperature and
moisture content, which are assumed to be the parameters that affect oil quality most
strongly. None of the samples comply with the PPO norm for rapeseed in Appendix B. For
6.7% moisture acid value and phosphorous content are too high. The other values are in
range. At decreased moisture level of 2.14% phosphorous content reduces close to 15 as
required by the standard at the cost of acid value. At 13.3% moisture acid value and
water content are unacceptable and other values are within limits. The data in Table 5-3
does not provide proper trends as most parameters decrease between 2.14-6.7% and
then decrease form 6.7%-13.3% or vice versa. Acid value tends to increase with
moisture content. Some possible explanations are given here. The strong decrease in
phosphorous content at 13.3% is probably due to lower temperature. Oxidation stability
is increased by increased water content. The higher carbon residue for 13.3% moisture
might be related to the darker colour of the oil due to dissolved pigments.

Table 5-3 Oil quality analysis performed by ASG
Parameter Unit Method Rapeseed standard 2.14% moisture 6.7% moisture 13.3% moisture
Acid value [mg KOH/g] DIN EN 14104 2.0 3.5 3.1 25.8
Phosphorous content [mg/kg] DIN EN 14107 15 16.2 23.2 2.4
Water content [mg/kg] DIN EN ISO 12937 750 741 733 1622
Oxidation stability 110°C [h] DIN EN 14112 5.0 12.6 10.4 21.3
Carbon residue [%] DIN EN ISO 10370 0.4 0.2 0.2 0.4
Processing oil temperature [°C] 97 85 52

5.10 Comparison of suitability for Jatropha processing
between BT50 and Sayari
Based on the results presented in this chapter higher oil recovery is achieved in the
Sayari expeller. At high residual oil content of around 12-15% repressing of the press
cake is possible, resulting in high oil recovery. The BT50 reaches higher oil recovery after
a single pass. However due to the lower residual oil content and the different press
geometry dual pass is not possible for Jatropha in the BT50. For research purposes the
BT50 is preferable because of its small capacity. Batches of around two kg are sufficient
for stable operation and relatively easy to condition. Severe jamming problems were
however experienced when operating at restriction sizes below 9mm and moisture levels
below the natural 6.7%. At 2.14% moisture level, the point at which oil recovery was
highest, continuous processing is not possible in the BT50 due to overheating and
jamming. The Sayari expeller is less suited for experimental use but operates much
better for full time large capacity production. When properly operated jamming of the
press can be avoided whereas this is not possible in most cylinder-hole presses like the
BT50.
5.10.1 Test conducted by industrial press manufacturers
To gain additional knowledge on the pressing process the two German press
manufacturers Keller (KEK EGON Keller GMBH & Co, Remscheid, Germany) and Reinartz
(Maschinenfabrik Reinartz GmbH & Co. KG , Neuss, Germany) were contacted for press
tests with Jatropha seeds. Testing results are shown in Table 5-4. More details on the
presses used for testing are given in Appendix L together with a table showing industrial
oil recovery rates for oilseeds. Rienartz additionally reported seed input temperature of
20°C and oil temperature between 30 and 60 °C increasing from feed section towards the
Chapter 5: Results and discussion


37
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
cake discharge. Oil recovery of these machines is higher than for the BT50 and similar to
the Sayari press. The advantage of the Keller and Reinartz presses compared to the
Sayari expeller is that they require only single pass of the material to reach a high oil
recovery.

Table 5-4 Results from tests at KEK Keller and Reinartz conducted in 2007. * [% oil/oil] means the
percentage of the oil present in the seeds that has been obtained by pressing
KEK Keller p0101 date sample description moisture oil contentpress cake oil yield [oil/oil]*
1 4/6/2007 standard settings start sample 6.7% 14.3% 72%
2 4/6/2007 standard settings end sample 6.7% 11.0% 79%
3 6/7/2007 changed pressure cones 6.7% 6.0% 89%
Reinartz AP08
3/7/2007 standard settings 6.7% 5-6.5% 88.3-91.1%

5.11 Conclusions
The most important findings based on the practical research conducted for this study is
presented here. Seed moisture content shows the strongest effect on oil recovery from
Jatropha. The natural moisture content of the seeds (6.7 wt.% w.b.) appears to be close
to optimal. Depending on the desired oil recovery pre-treatments might therefore not be
necessary. Restriction size and rotational speed of the screw are two other influential
parameters. Restriction size should be reduced as far as possible taking into
consideration both operational limits of the press and the effect of temperature on oil
quality. Lower rotational speeds lead to higher oil recovery, which is mainly attributed to
higher residence time. Speed can only be reduced to a limited extent as below 20 RPM
operation is not possible. The choice of rotational speed varies with the demands of the
user. For high production capacity rotational speed should be high and for high oil
recovery per kg input material it should be low. Decrease in hull content below a
threshold value between 66-80% has a strong, although negative, effect on oil recovery.
The influence of seed pre-heating and flaking/crushing proved to be insignificant for
Jatropha seeds. A pre-treatment of one hour cooking followed by seed drying results in
the highest oil recovery for both presses. Single pass pressing of cooked and dried seeds
resulted in the highest measured oil recovery of 89% for both BT50 and Sayari press.
After dual pass oil recovery for the Sayari increased to 91%, being close to the theoretic
maximum of 95%. For comparison normal seeds yields were 79% for the BT50 (single
pass) and 88% for the Sayari (dual pass). Single pass pressing of cooked seeds in the
Sayari expeller thus leads to higher yields than normal seeds after dual pass pressing.
This pre-treatment can thus be used to decrease energy consumption and increase
processing capacity. The energy requirement is of subordinate importance as for the
BT50 the energy input per litre oil varies between 1% and 2.5% of the calorific value of
Jatropha oil.

Taking into consideration all the test results optimal oil recovery is expected at a
moisture content of 2-4% after cooking at 70°C, 100% hull content and the smallest
restriction size and lowest speed possible for a certain press type. Unfortunately the oil
yield using this setup could not be determined due to unexpected breakdown of the press
in the final testing stages. To summarize an overview of the effects of pressing
parameters on the efficiency and conditions inside the press barrel is visible in Table 5-5.

Chapter 5: Results and discussion


38
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Table 5-5 Overview of the effects of press settings and seed pre-treatment on oil recovery,
pressure, temperature, throughput and energy use
Oil recovery Pressure Temperature Throughput Energy/liter
Press parameters
RPM -
restriction size -
Seed treatments
heating - -
flaking - -
moisture content
hull fraction
boiling -
Chapter 6: Possibilities for oil pressing in rural Tanzania


39
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
6 Possibilities for oil pressing in rural Tanzania
This chapter puts the results on Jatropha expression from previous chapters into a
broader context. Some of the experimental results were used to make recommendations
on how to set up Jatropha oil production in rural Tanzania.
6.1 Introduction
Occasion
Over the last decades the western approach of setting up development projects in less
developed countries has proven extremely ineffective (Douthwaite, 2002). Is Jatropha
just another opportunistic outburst trying to find a solution to global poverty problems or
is it really possible to fight poverty using Jatropha? Despite all euphoria concerning
Jatropha knowledge on how to actually process it into oil as to make it profitable for rural
communities is fairly limited.

Goal
To see whether Jatropha has potential to improve the situation in poor tropical countries
one should think of how to setup local production. The main goal of this chapter is
therefore to find out whether it is preferable to centralize or decentralize Jatropha oil
production when decentralized cultivation of Jatropha is assumed? Furthermore
calculations are done to determine if Jatropha can be more than just a primary
agricultural product in rural areas in less developed countries

Approach
For local communities to financially benefit from Jatropha cultivation probably requires
cooperation between investors and small farmers. Centralized and decentralized
production of Jatropha oil are compared using the cost benefit analysis (CBA) method. As
an extension another CBA was conducted to judge the viability of oil production on village
level. However economic viability in itself is not sufficient to establish successful projects
thus some local circumstances are taken into consideration before making
recommendations.

Data on the costs and benefits of small scale Jatropha oil processing for rural
communities was gathered during a visit in Tanzania from January to April 2007. Most of
the data was obtained by interviewing, taking questionnaires and the writers
interpretation of the situation at hand. Prices of machinery were obtained from the
German screw press manufacturers Reinartz and Keller. Prices for transport and labour
are estimates based on Tanzanian prices.

Structure
The setup of the rest of this chapter is as follows. Paragraph 6.2 shortly explains CBA
theory. Research outcomes are presented and discussed in paragraph 6.3. Subsequently
paragraph 6.4 passes judgement on the viability of Jatropha oil production on village
level. Conclusions are drawn in paragraph 6.5.
6.2 Cost Benefit Analysis: Theory
The CBA is designed as a tool to help investors to select projects that will contribute most
to their objectives while keeping in mind their limited financial resources (Romijn and
Biemond, 2005). This method has been developed by the World Bank and the United
Nations Industrial Development Organisation (UNIDO) and is often referred to as project
appraisal. Although one can distinguish between a financial, economic, social and
environmental CBA this report will only focus on the financial aspect as this is relevant for
the question at hand and fits well to the available data.

Chapter 6: Possibilities for oil pressing in rural Tanzania


40
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
In a CBA the distinction between cost and expenditures, and between income and
revenues is not relevant. As a CBA takes the entire life-time of a project as its time
horizon total costs will be equal to total expenditures by definition (Romijn and Biemond,
2005). The same holds for benefits and income. This is best explained looking at
depreciation. Depreciation is not taken into account in a CBA as it leads to costs without
leading to expenditures. Normally a CBA consists of the following three steps:

1. Profitability assessment
2. Liquidity assessment and determination of a financial plan
3. Sensitivity analysis
6.2.1 Profitability assessment
When judging project profitability the way of financing determines the interest rate to be
used to judge the investments profitability. For borrowed money the interest rate on
loans is applied while for own money the interest rate on savings is used. The core of the
CBA analysis is to compare the earnings profile of the project to the going interest rate
which represents the costs of financing as discussed above (Romijn and Biemond, 2005).
The most frequently used criteria in this comparison are the Net Present Value (NPV) and
the Internal Rate of Return (IRR). Both criteria use discounting techniques which take
into account the effect of interest that would be received if the money was in the bank.
Multiplying non-financial cash flows by the discount factor yields discounted cash flows.

NPV
The NPV is the sum of the discounted non-financial cash flows during the projects
lifetime, thus providing the exact sum of money which a project is expected to generate.
For an NPV > 0 a project is expected to yield more than the interest rate. Therefore the
project is viable in principle although liquidity and risk assessment should be conducted.
A NPV =< 0 indicates no commercial reason to embark in the project.

IRR
The IRR is a measure for the extent of profitability of the project compared to the going
interest rate and is determined by setting the NPV to zero. For IRR > market interest rate
‘i’ the project is more profitable than the cost of financing the project. For IRR =< market
interest rate ‘i’ the yield is equal to or less than the cost of financing.

Inflation
Because of Tanzania’s significant inflation rate it is convenient to work with so called
constant prices. High inflation rates make it difficult to estimate prices of inputs 5 or 10
years from now. Constant prices are the prices in the base year of the CBA. The
advantage of using constant prices is that the same prices can be used without taking
inflation into account. The effects of inflation and interest are then combined in the
discount factor:
t
r
factor Discount
) 1 (
1
+
=
with t = number of years
r = real discount rate

1
) 1 (
) 1 (

+
+
=
p
i
r
with i (interest rate) = 15.54%
1

p (rate of inflation) = 6.9%
2


1
Source: website of the Bank Of Tanzania: http://www.bot-tz.org/publications/EconomicIndicators/
InterestRates.htm, visited 04-06-2007
2
Source: Website of the Bank Of Tanzania: http://www.bot-tz.org/publications/inflationDevelopments.htm,
visited 04-06-2007
Chapter 6: Possibilities for oil pressing in rural Tanzania


41
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

6.2.2 Liquidity assessment
In principle the liquidity assessment determines whether a project is able to fulfil
payment obligations from its cash inflows during its entire lifetime (Romijn and Biemond,
2005). As liquidity does not add anything to the question asked in this chapter it will not
be taken into further consideration.
6.2.3 Sensitivity analysis
As projects are always subjected to uncertainty one should assess the sensitivity of the
results of the CBA to changes in specific costs and benefits. This is done by executing a
sensitivity analyses in which only one variable at a time can be changed to see how this
affects NPV and IRR. By calculating so called switching values (the degree of change in a
value that sets NPV to zero) one can determine the risk of a project.
6.3 CBA results: comparison of centralized and decentralized
production
This paragraph focuses on the most viable production options for centralized and
decentralized production. The CBA was used to compare centralized and decentralized
processing of Jatropha seeds into oil and press cake. The difference between central and
dispersed positioning of pressing facilities is represented in Figure 6-1. Obviously the
distance along which input material has to be transported is considerable reduced in case
of decentralized processing. The main variables included in the CBA are investment cost
for equipment, seed purchasing price, labour costs, equipment maintenance costs and
revenues. Table 6-1 shows the input values used for calculations. As can be seen in the
table a comparison is made between three different processing configurations. More
details on the different presses is given in Appendix C, F and K. In case of decentralized
pressing the values represent expenses for all machines at multiple locations. The cases
for centralized pressing incorporate only one location and one machine. Prices are all
given in dollars as this is the foreign monetary unit used in Tanzania. In addition to
difference in press type and size different oil cleaning methods were compared. The
results for different types of presses and filtering methods can be found in Appendix M.
The values in Table 6-1 were computed with a combination of data sources, assumptions
and calculations. The main assumptions are listed below.

Main assumptions

• From previous chapters the oil recovery is set at 85%. Sediment level is 40 vol%
and contains 55 wt.% oil.

• Production capacity is a measure for the oil production. Supply of seeds is not
taken into account in this respect.

• Lifetime of German presses is 20 years, while Tanzania Sayari expeller needs to
be replaced after 5-10 years. In CBA calculations write-off in 5 years is assumed
for the Sayari.

• The selling price for Jatopha oil is 750 TZS/l which is equivalent to $0.59. This
price is reasonable taking into account that fuel taxes have not yet been included
in the calculation.


• Cost price of an 8 ton capacity truck is $15,000 based on a small Mercedes truck
and reducing the price to Tanzanian levels. For increased production capacity the
amount of kilometres a truck cover per year remain the same. Instead the
number of trucks needed changes with travel distance and production capacity.
Chapter 6: Possibilities for oil pressing in rural Tanzania


42
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Therefore write-off is 10% in all cases. 3, 6 and 12 trucks are assumed for
decentralized, centralized and large scale centralized 24h/day production
respectively.

• Press cake is sold for $0.05/kg without further processing.

• One operator is required for each press independent of its size.

• Two seed collectors are assumed to be on each truck.











Figure 6-1 Area division for centralized and decentralized production. The 5 dotted circles
indicate the areas for decentralized production with their own pressing facility. The large circle
represents centralized pressing which is located at the hatched circle.

Chapter 6: Possibilities for oil pressing in rural Tanzania


43
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Table 6-1 Input values for the cost benefit analysis
Monetary Reference
exchange rate (dollar to chilling) 1268 www.bot-tz.org
discount factor 20.31% www.bot-tz.org
interest rate 15.54% www.bot-tz.org
inflation rate 6.90% www.bot-tz.org
real discount rate 8.08%
Press type 5 x KEK P0101 1 x Reinartz AP14/30 1 x Reinartz AP15/45 Data Source
(decentralized) (centralized) (centralized)
Investment
press $126,965 $87,125 $101,243 Reinartz & KEK Egon Keller
cyclone filter $232,704 $101,914 $140,411 Reinartz
storage 5000L 8000L 14000L
$115,765 $26,304 $32,606 Reinartz
Maintenance
annual maintenance (5%) $6,348 $4,356 $5,062
capacity (kg Jatropha/hr) 350 350 700 Reinartz & KEK Egon Keller
annual capacity ton (8 h operation) 658 658 1,316
annual capacity ton (24 h operation) 2,764 2,764 5,527
Seed buying costs
price/kg $0.09 $0.09 $0.09 Diligent Tanzania LTD
8h $62,271 $62,271 $124,543
24h $261,539 $261,539 $523,079
Fuel use for pressing
Annual fuel use 8h (liter) 9,400 5,640 11,280 Reinartz & KEK Egon Keller
Annual fuel use 24h (liter) 39,480 23,688 47,376 Reinartz & KEK Egon Keller
Amount usable oil
8h (liter) 205,165 208,925 417,850
24h (liter) 861,694 877,486 1,754,972
Fuel costs
capacity (ton/truck) 8h 8 8 8 Mercedes 815 Atego
number of rides 8h 82 82 165
average km/ride 300 750 750
total km 8h 24,675 61,688 123,375
km/liter 6 6 6 Mercedes 815 Atego
price /liter $0.99 $0.99 $0.99
total fuel 8h $4,071 $10,178 $20,357
total fuel 24h $12,214 $30,535 $61,071
truck price $15,000 $15,000 $15,000 Mercedes 815 Atego
Labour cost
wage dollar/month $80.00 $80.00 $80.00
wage /hr $0.45 $0.45 $0.45
8h
press operator $5,054 $842 $842 Diligent Tanzania LTD
seed collectors $1,893 $11,356 $22,713 Diligent Tanzania LTD
manager $41,100 $41,100 $41,100
total $48,047 $53,299 $64,655
24h
press operator $21,229 $3,538 $3,538 Diligent Tanzania LTD
seed collectors $5,678 $34,069 $68,139 Diligent Tanzania LTD
manager $82,200 $82,200 $82,200
total $109,107 $119,808 $153,877
Revenues
oil 8h $121,352 $123,576 $247,151
oil 24h $509,677 $519,018 $1,038,035
presscake 8h $23,030 $23,030 $46,060
prescake 24h $96,726 $96,726 $193,452


6.3.1 Decentralized pressing
Decentralized processing in all cases showed negative NPV at 8 hour operation per day as
can be seen in Appendix M. Centralized processing was possible at 8 hour operation
although 24 hours was also strongly preferred. The values in Table 6-2 to Table 6-8 are
best guesses. Rounded numbers would only increase inaccuracy. Table 6-2 shows a
positive NPV for decentralized processing during 24 hours a day. When comparing the
real IRR of 34% to the real interest rate of 8.08% shows this option is quite profitable.
The positive value for equipment in year 5 represents the NPV of the equipment taking
into consideration the difference in write of period between German and Tanzanian
Chapter 6: Possibilities for oil pressing in rural Tanzania


44
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
machines. Write-off periods of 5, 10 and 20 years are taken for respectively Tanzanian
machines, trucks and German machines.

Table 6-2 Calculation of NPV and IRR for decentralized oil production capacity of 350kg/hr
5 x KEK P0101 (70kg/hr) 24 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$567,977 -$567,977 1.000 -$567,977
1 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.925 $197,227
2 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.856 $182,482
3 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.792 $168,840
4 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.733 $156,218
5 $379,076 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $592,238 0.678 $401,578
NPV $538,367
IRR 34%


Investment cost can be reduced by selecting local technologies like the Sayari expeller.
Table 6-3 shows a significant increase in both NPV and IRR compared to using the more
expensive German technology.

Table 6-3 Calculation of NPV and IRR for decentralized oil production capacity of 350kg/hr using the
cheap Sayari expeller
5 x Sayari (70kg/hr) 24 hours operation per day with cyclone filter and cheaper storage
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$329,575 -$329,575 1.000 -$329,575
1 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.925 $198,357
2 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.856 $183,528
3 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.792 $169,807
4 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.733 $157,113
5 $197,029 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $411,413 0.678 $278,966
NPV $658,196
IRR 63%

6.3.2 Centralized pressing
Centralized production at equal production capacity results in a higher NPV due to
reduced investment costs. As can be seen from the IRR of 61% in Table 6-4 profitability
of the project increases significantly when production is centralized compared to the
situation in Table 6-2.

Table 6-4 Calculation of NPV and IRR for centralized oil production capacity of 350 kg/hr
1 x Reinartz AP 14/30 (350kg/hr) 24 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$310,760 -$310,760 1.000 -$310,760
1 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.925 $181,915
2 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.856 $168,315
3 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.792 $155,732
4 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.733 $144,090
5 $195,518 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $392,132 0.678 $265,892
NPV $605,185
IRR 61%

Judging from Table 6-5 economies of scale seem to have more effect on profitability than
the way production is organized. At doubled production the IRR increases from 61% to
99% making the much more profitable. Comparing Table 6-4 and Table 6-5 five years
NPV increased with over $1,000,000 for a $170,000 increase in initial investment.

Table 6-5 Calculation of NPV and IRR for centralized oil production capacity of 750 kg/hr
1 x Reinartz AP15/45 (750kg/hr) 24 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$481,686 -$481,686 1.000 -$481,686
1 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.925 $447,203
2 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.856 $413,770
3 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.792 $382,837
4 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.733 $354,216
5 $271,240 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $754,577 0.678 $511,654
NPV $1,627,994
IRR 99%

Chapter 6: Possibilities for oil pressing in rural Tanzania


45
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
6.3.3 Sensitivity analysis
A sensitivity analysis (see table 6-6 to 6-8) of the different cases shows centralized
production is less sensitive to fluctuating values than decentralized production. In
addition Table 6-8 shows sensitivity decreases for larger production scale, which is to be
expected. The project shows very little sensitivity to changes in most variables. Changes
in raw material price and oil selling price and revenues show the biggest impact. The
extremely high switching value for fuel costs in all three cases suggests underestimation
of travel distances and fuel consumption. Normally transport costs are expected to
contribute to a larger extent. Revenues have the lowest switching variable which still has
a margin of >20% in all cases. For 24hr/day operation the project shows little risk for
both centralized and decentralized production according to the sensitivity analysis and
can therefore be implemented in any of the three setups. Clearly the larger the
production scale the less sensitive the process is to the changes taken into account in this
paragraph.

Table 6-6 Sensitivity analysis decentralized production

Sensitivity analyses 24h operation KEK P0101
Cash flow item Assume Effect IRR Effect NPV Switching value
Equipment costs 5% increase -2% -$30,000 + 95%
Seed costs 25% increase -12% -$260,000 +52%
Fuel costs 10% increase 0% -$4,000 +1116%
Labour costs 10% increase -2% $40,000 +129%
Revenues 10% decrease -11% -$240,000 -23%
Jatropha oil price 20% increase +29% $400,000 -27%


Table 6-7 Sensitivity analysis centralized production
Sensitivity analyses 24h operation Reinartz AP14/30
Cash flow item Assume Effect IRR Effect NPV Switching value
Equipment costs 5% increase -3% -$15,000 +195%
Seed costs 25% increase -22% -$260,000 +58%
Fuel costs 10% increase -1% -$12,000 +497%
Labour costs 10% increase -4% -$50,000 +127%
Revenues 10% decrease -21% -$245,000 -25%
Jatropha oil price 20% increase +34% 400,000 -29%


Table 6-8 Sensitivity analysis centralized production
Sensitivity analyses 24h operation Reinartz AP15/45
Cash flow item Assume Effect IRR Effect NPV Switching value
Equipment costs 5% increase -5% -$25,000 +338%
Seed costs 25% increase -26% -$520,000 +78%
Fuel costs 10% increase -1% -$25,000 +669%
Labour costs 10% increase -3% -$60,000 +266%
Revenues 10% decrease -26% -$490,000 -33%
Jatropha oil price 20% increase 54% $830,000 -39%


6.3.4 Non financial consideration
Determination of the optimal pressing arrangement should not solely be based on
financial considerations. Some additional factors that should be taken into account during
decision making are listed here:

• The oil recovery for large presses is only slightly higher than for smaller presses
because the design is often very similar. Table L-3 in Appendix L shows that the
differences are expected to be within 1-2% for most oilseeds

• When production is centralized large presses should be used to reduce investment
costs. Local people in less developed countries are generally unfamiliar with these
Chapter 6: Possibilities for oil pressing in rural Tanzania


46
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
large and complicated machines. In case of centralized pressing operational
mistakes could cause the whole production to come to a halt or cause expensive
machine damage. For decentralized pressing 80% of the production capacity
remains in case of a single breakdown and repair cost will be lower.

• Small decentralized presses allow more flexibility in the expansion of production
capacity. Investments costs are relatively low and capacity can be gradually
increased.

• Expansion into new geographic areas is easier with decentralized production as
fewer investments are required to setup new pressing units at other locations.

• In case of problems Tanzanian people prefer personal visits over telephone or
email. Scattered production locations reduce the distance between farmer and oil
producer and can therefore improve communication between both parties. This
could in turn lead to improved product quality and yield as pressing locations
could function as knowledge centres.

Incorporating some points mentioned in the ‘learning selection model’ from Douthwaite
(Douthwaite, 2002) gives rise to some additional comments:

• Judging from experiences during field trips in Tanzania local farmers seemed
moderately motivated to grow Jatropha or collect seeds from wild Jatropha trees.
The fact that most trees were not harvested in areas where Diligent was actively
buying suggest lack of motivation on the side of the farmer. This might be caused
by low prices or the inability of local farmers to judge the long term benefits.

• Although locals are perhaps unable to technologically modify the screw press they
are able to come up with new applications. An example is the use of a multi
functional platform for pressing, milling and electricity generation. Linking
Jatropha to such directly noticeable benefits can increase motivation in local
communities.

• According to Douthwaite the need of an innovation should be great. This issue is
questionable fro Jatropha oil. Most remote villages have no direct application for
Jatropha oil and to them Jatropha is just another cash crop. There might of course
be a need for the advantages of using Jatopha for electricity or cooking, but one
should take care not to confuse our needs with the needs of the local population.

• The last point mentioned here is the necessity of a product champion. This is
someone who is respected in his community and who works with users and
producers to identify and solve problems in order to increase the fitness of a
technology. The innovation champion will also promote the beneficial applications
of the new technology. With respect to Jatropha the activities of a product
champion could be to motivate farmers to grow Jatropha and to make them aware
of long term benefits of increased income security. The role of product champion
could for example be filled by a village leader.

This short consideration of innovation theory was restricted to some of the significant
success factors. However the innovation theory discussed by Douthwaite contains many
other factors that should be satisfied.
6.4 Viability of community level processing: Engaruka case
The village of Engaruka consists of two separate parts: northern and southern Engaruka.
Most of the 9000 inhabitants of this Maasai village make a living as farmers or keeping
feedstock. As is the case for the village Temi ya Simba there is no grid connection
present in the village. Therefore a Multi Functional Platform (MFP) was installed by a
Chapter 6: Possibilities for oil pressing in rural Tanzania


47
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Tanzanian NGO named TaTedo in November 2006 to combine seed pressing with
electricity generation and grain milling. The MFP operates five hours a day generating
roughly 6 kWh. In total 20 shops, 3 houses and a campsite are provided with energy
from this platform. In addition to the grain mill connected to the MFP seven more diesel
powered grain mills are present in the village. Together with the diesel driven generator
these machines constitute the potential Jatropha consumption in Engaruka. As Jatropha
was already grown in Engaruka for fencing most seeds still stem from hedges. Over 30
farmers are growing Jatropha on small plots between 0.5-3 acres in size (Zheng and
others, 2005, 193-202). The first harvest in 2007 produced approximately 2400 kg
Jatropha seeds. The total land area available for Jatropha is around 100 acres. Figure 6-2
shows a model for Jatropha utilization in Engaruka. A press cooperation consisting of
maybe 5-10 people could start production of Jatropha oil and grain millers could also
benefit from this cheaper locally produced fuel to power their engines. The investment
required to setup a press-cooperation probably exceed the amount of money available to
the local investors. Most of the villagers cannot get any loans from the bank as farming is
excluded from the definition of enterprises the bank uses when judging loan requests.
Therefore Savings and Credit Co-operatives (SACCOs) are included in this system. The
SACCO system is a mutual membership organization, which involves pooling of voluntary
savings from members in the form of shares. These savings or shares form the basis for
providing credit to members. The available credit is normally around three times the level
of savings or shares. The importance of this market segment is underlined by a SACCOs
market of nearly US$1 billion in Kenya.


Figure 6-2 Jatropha utilization in Engaruka

The possibility of village level processing of Jatropha seeds into oil was evaluated using a
method similar to the previous paragraph. Calculations shown in Appendix N show that
although village level pressing is profitable larger profits are achieved by just selling
seeds. Assumptions in this comparison are:

• 100 acres is available for growing Jatropha without decreasing food production.

• In case of local processing Jatropha oil is sold at $0.59 and press cake is used as
fertilizer.

Chapter 6: Possibilities for oil pressing in rural Tanzania


48
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
• When only selling seeds investment costs are assumed to be zero and fertilizer
costs are equal to the value of the fertilizer produced in case of processing seeds
to oil ($0.05 /kg).

• Labour costs are zero as farmers grow Jatropha for additional income besides their
normal activities.

• Calculations for Jatropha pressing assume a small village cooperation that uses an
MFP to press seeds, produce electricity and mill maize. The availability of oil and
electricity are of additional value to the whole village.

• Calculations for selling seed include no local electricity and oil production.

Table 6-9 NPV and IRR for various situations in Engaruka
50 acres low yield 100 acres low yield 100 acres good yield 100 acres low yield seeds 100 acres good yield seeds
NPV $2,950 $14,500 $51,500 $12,750 $54,500
IRR 33% 115% 353% - -


Judging from Table 6-9 pressing on village level can contribute to the welfare level in
Engaruka, especially when taking into account the increase in welfare due to availability
of oil and electricity. For good yields the additional income can amount up to one hundred
annual wages. In case of 100 acres and low yield the increased revenues are equivalent
to 19 monthly salaries. As was the case when comparing centralized and decentralized
processing in the previous paragraphs larger production means higher returns (see Table
6-9).

Initially two other villages were reviewed on Jartopha application. Jatropha use in these
villages was not as extensive as in Engaruka. One of the villages only sells seeds and in
the other limited quantities of oil could be pressed to fuel cooking stoves. Because of its
small scale village level production of Jatropha for cooking is expected to be inefficient
taking into account the modest profits in the much larger Engaruka case. The CBA for
Engaruka suggests that Jatropha activities on a scale smaller than this should be
restricted to direct selling the primary product. More information on all three villages can
be found in Appendix O.
6.5 Conclusions
One should keep in mind that the CBA explained in this chapter provides insight in the
profitability of a project. This method was used to select the best option from a range of
possibilities. Calculated NPVs are no indication for the absolute value of commercial
profits. The initial investment costs for Jatropha oil production are high compared to
expected revenues. Daily production time of 8 hours is expected not to be profitable
based on the executed CBA. When production is running 24 hours a day, 7 days a week
production showed profitable. Based on purely financial results from the CBA large scale
centralized processing is preferred over decentralized pressing. For similar production
scale centralization allows an IRR of 61% compared to 30-40% in case of
decentralization. However, reviewing both financial and non-financial aspects shifts
preference to decentralized pressing for Tanzania and similar countries. In case of non-
commercial projects the use of local technology is desirable as this creates financial
activity and potentially yields higher revenues. Commercial firms might prefer more
robust technology, which is much more reliable and durable at the cost of higher initial
investments. For projects to become successful care should be taken to bear in mind the
importance of local circumstances such as motivation, need of a technology and the
presence of a product champion.

Jatropha has significant added value for rural communities in Tanzania. For a village of
9000 inhabitants and 100 acres of available land the NPV of Jatropha cultivation appears
to be between $12,750 and $54,500 per year depending on seed yields, which is
Chapter 6: Possibilities for oil pressing in rural Tanzania


49
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
equivalent to approximately 100 annual minimum wages. Oil processing on village scale
without the intervention of commercial firms seems preferable to mere seed selling if
coupled to electricity generation and maize milling. Oil processing for 100 acres of
Jatropha could easily provide an additional $1,500 per annum, which is equivalent to 19
monthly salaries to be divided between the initiators. With this extra income comes the
availability of electricity and cheap fuel. In depth investigation of financing possibilities
through micro-credit is advisable.
Chapter 7: Conclusions and recommendations


50
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
7 Conclusions and recommendations
The main question to be answered by this study is as follows:

• How can oil recovery, energy input and cost of Jatropha seed expression in screw
presses be optimized while retaining a quality that complies with international fuel
standards?

Three main options to increase oil recovery have been identified: pre-treatment of seeds
before pressing, changes in press operation and changes in press design. Only the first
two could be included in the present study, although two different press types were
included in this study. Both seed pre-treatment and press operation showed significant
impact in oil recovery and energy requirement. Quality was mostly affected by moisture
content and temperature, the latter of which can be due to both seed pre-treatment and
press operation. None of the samples tested for quality met standards for Pure Plant Oil.
However standard treatments are likely to bring the quality parameters within the range
of these standards. Seed moisture content (varied from 2-13%) shows the strongest
effect on oil recovery from Jatropha. Oil yield for different moisture levels varies from 33-
88%. The natural moisture content of the seeds (6.7 wt.% w.b.) appears to be close to
optimal with an oil yield of 79%. Depending on the desired oil recovery pre-treatments
might therefore not be necessary. Restriction size and rotational speed of the screw are
two other influential parameters. If high yield are required restriction size should be
reduced as far as possible taking into consideration both operational limits of the press
and the effect of temperature on oil quality. Lower rotational speeds lead to higher oil
recovery, which is mainly attributed to longer residence time. Speed can only be reduced
to a limited extent as below 20 RPM operation is not possible with the BT50. The choice
of rotational speed varies with the demands of the user. For high production capacity
rotational speed should be high and for high oil recovery per kg input material it should
be low. Oil yields drastically decrease below 80% hull content, from 80-100% the yield is
hardly influenced. The influence of seed pre-heating and flaking/crushing proved to be
insignificant for Jatropha seeds. A pre-treatment of one hour cooking followed by seed
drying results in the highest oil recovery for both presses. Single pass pressing of cooked
and dried seeds resulted in the highest measured oil recovery of 89% for both BT50 and
Sayari press. After dual pass oil recovery for the Sayari increased to 91%, being close to
the theoretic maximum of 95%. For comparison normal seeds yields were 79% for the
BT50 (single pass) and 88% for the Sayari (dual pass). Single pass pressing of cooked
seeds in the Sayari expeller thus leads to higher yields than normal seeds after dual pass
pressing. This pre-treatment can thus be used to decrease energy consumption for
pressing and increase processing capacity. The energy requirement is of subordinate
importance as for the BT50 the energy input per litre oil varies between 1% and 2.5% of
the calorific value of Jatropha oil. For all tests the amount of foots (30-50 vol.%) in the
crude oil was problematic for filtering.

Taking into consideration all the test results optimal oil recovery is expected at a
moisture content of 2-4% after cooking at 70°C, 100% hull content and the smallest
restriction size and lowest speed possible for a certain press type.

Additional consideration of the practical application of screw presses for the production of
Jatropha oil in Tanzania led to answers to the following two additional questions.

• What is the optimal setup for the production of Jatropha oil in Tanzania?

• In what way can rural areas in which their use is anticipated benefit from Jatropha
cultivation?

Chapter 7: Conclusions and recommendations


51
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
A ‘Cost Benefit Analysis’ (CBA) was used to answer both questions. One should keep in
mind that the CBA explained in this chapter provides insight in the profitability/viability of
a project. Calculated ‘Net Present Values’ (NPVs) are no indication for the absolute value
of commercial profits. The CBA method was used to select the best option from a range
of possibilities (decentralized, central and high capacity central). The initial investment
costs for Jatropha oil production are high compared to expected revenues. Daily
production time of 8 hours is expected not to be profitable based on the executed CBA.
When production is running 24 hours a day, 7 days a week production showed profitable.
Based on purely financial results from the CBA large scale centralized processing is
preferred over decentralized pressing. For similar production scale centralization allows
an ‘Internal Rate of Return’ IRR of 61% compared to 30-40% in case of decentralization.
However, reviewing both financial and non-financial aspects shifts preference to
decentralized pressing for Tanzania and similar countries. In case of non-commercial
projects the use of local technology is desirable as this creates financial activity and
potentially yields higher revenues. Commercial firms might prefer more robust
technology, which is much more reliable and durable at the cost of higher initial
investments.

Jatropha appears to have significant added value for rural communities in Tanzania. For a
village of 9000 inhabitants and 100 acres of available land the NPV of Jatropha cultivation
appears to be between $12,750 and $54,500 per year depending on seed yields, which is
equivalent to approximately 100 annual minimum wages. Oil processing on village scale
without the intervention of commercial firms seems preferable to mere seed selling if
coupled to electricity generation and maize milling. Oil processing for 100 acres of
Jatropha could easily provide an additional $1,500 per annum, which is equivalent to 19
monthly salaries to be divided between the initiators. With this extra income comes the
availability of electricity and cheap fuel in rural areas.

Some recommendations for further research are listed below:

• This study was only related to screw pressing process, which is only part of the
production process of Jatropha oil. Additional research aimed at designing the
total production process would provide better insight on process efficiencies and
viability. Solutions for filtering the crude oil could for example result in higher
yields of clean oil than optimizing the oil recovery yield of a screw press.

• Optimization of industrial presses for Jatropha seeds might still be interesting to
recover a few additional percentages of oil and reduce the amount of foot in the
crude oil.

• Decentralized setup of Jatopha press plants is preferred over centralized
processing although CBA analysis predicts higher profits in case of centralization.
The main reason is higher flexibility to adapt to various local circumstances.

• Repetition of press tests and measurements on oil recovery would increase the
reliability of the results.

• In depth investigation of financing possibilities through micro-credit is advisable
with a view to planning community Jatropha projects in Tanzania. A closer look at
the success factor given by Douthwaite is advisable when actually starting a
project for Jatropha processing in any developing country.

Reference list


52
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Reference List

Jatropha Handbook. FACT FOUNDATION. 2006.
Ref Type: Unpublished Work
Interview with Bruno S. Rubaza; seed collector Temi ya Simba for Diligent Tanzania in.
2007.
Ref Type: Personal Communication
Aderibigbe, A. O., Johnson, C. O. L. E., Makkar, H. P. S., Becker, K., and Foidl, N., 1997,
Chemical composition and effect of heat on organic matter- and nitrogen-degradability
and some antinutritional components of Jatropha meal: Animal Feed Science and
Technology, v. 67, p. 223-243.
Agarwal, D. and Agarwal, A. K., 2007, Performance and emissions characteristics of
Jatropha oil (preheated and blends) in a direct injection compression ignition engine:
Applied Thermal Engineering, v. 27, p. 2314-2323.
Akintayo, E. T., 2004, Characteristics and composition of Parkia biglobbossa and Jatropha
curcas oils and cakes: Bioresource Technology, v. 92, p. 307-310.
Augustus, G. D. P. S., Jayabalan, M., and Seiler, G. J., 2002, Evaluation and bioinduction
of energy components of Jatropha curcas: Biomass and Bioenergy, v. 23, p. 161-164.
Bargale, P. C. and Singh, J., 2000, Oil expression characteristics of rapeseed for a small
capacity screw press: Journal of food science technology, v. 37, p. 130-134.
Benge, M. Assessment of the potential of JAtropha Curcas (biodiesel tree) for energy
production and other uses in developing countries. 2006.
Ref Type: Report
Biswas, S., Kaushik, N., and Srikanth, G. Biodiesel Conference Towards Energy
Independence - Focus on Jatropha:
Biodiesel: Technology and Business Opportunities ñ An Insight. Brahma Singh, Maj Gen,
Swaminathan, R., and Ponraj, V. 10-6-2006. Technology Information, Forecasting and
Assessment Council (TIFAC) Department of Science and Technology, New Delhi.
Ref Type: Conference Proceeding
DGF-Einheitsmethoden, 2002, Deutsche einheitsmethoden zur untersuchung van fetten,
fettprodukten, tensiden und verwandten stoffen: stuttgart, Wissenshaftliche
Verlagsgesellschaft mbH.
Douthwaite, B., 2002, Enabling Innovation: Apractical guide to understanding and
fostering technological change: London, Zed Booka Ltd.
Dufaure, C., Leyris, J., Rigal, L., and Mouloungui, Z., 1999, A Twin-screw extruder for oil
extraction: direct expression of oleic sunflower seeds: Journal of the American Oil
Chemists' Society, v. 76, p. 1073-1080.
Eggers, R., Broeck, H., and Stein, W., 2006, Messung von Druck- und Temperaturprofilen
beim abpressen van Ölsaaten in Seiherschneckenpressen: Fette Seifen Anstrichmittel, v.
87, p. 494-499.
Reference list


53
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Enterprising Solutions Global Consulting, LLC. UNCDF microfinance programme impact
assesment 2003. 2003.
Ref Type: Report
Faborode, M. O. and Favier, J. F., 1996, Identification and Significance of the Oil-point in
Seed-oil Expression: Journal of Agricultural Engineering Research, v. 65, p. 335-345.
Ferchau, E. Equipment for decentralized cold pressing of oil seeds. 1-11-2000.
Folkecenter for Renewable Energy.
Ref Type: Report
Foidl, N. Isopropanol injection in extruder. 2007.
Ref Type: Personal Communication
Gubitz, G. M., Mittelbach, M., and Trabi, M., 1999, Exploitation of the tropical oil seed
plant Jatropha curcas L: Bioresource Technology, v. 67, p. 73-82.
Gupta, R. K. and Das, S. K., 2000, Fracture resistance of sunflower seed and kernel to
compressive loading: Journal of food engineering, v. 46, p. 1-8.
Haas, W. and Mittelbach, M., 2000, Detoxification experiments with the seed oil from
Jatropha curcas L: Industrial Crops and Products, v. 12, p. 111-118.
Hartlieb Euler and David Gorriz. Case Study "Jatropha Curcas". 2004. Global Facilitation
Unit for Underutilized Species (GFU) & Deutsche Gesellschaft für Technische
Zusammenarbeit (GTZ) GmbH.
Ref Type: Report
Heller, J. Physic nut Jatropha Curcas L.; promoting the conservation and use of
underutilized and neglected crops 1. 1996. Institute of Plant Genetics and Crop Plant
Research, Gatersleben/ International Plant Genetic Resources Institute, Rome.
Ref Type: Serial (Book,Monograph)
Henning, Reinhard K. The Jatropha System - economy and dissemination strategy;
Integrated Rural Development by Utilisation of Jatropha curcas L. (JCL)
as Raw Material and as Renewable Energy. 6-4-2004. Bagani.
Ref Type: Conference Proceeding
Henning, Reinhard K. Jatropha Curcas L. in Afrika: Assessment of the impact of the
dissemination of "the Jatropha System" on the ecology of the rural area and the social
and economic situation of the rural population (target group) in selected countries in
Africa. 2007. Bagani.
Ref Type: Report
Hui, Y. H. Bailey´s industrial oil & fat products volume 2:" Edible oil &fat products: oil
and oil seeds". Hui, Y. H. 1996. John Wiley & Sons.
Ref Type: Report
Kandpal, J. B. and Madan, M., 1995, Jatropha curcus: a renewable source of energy for
meeting future energy needs: Renewable Energy, v. 6, p. 159-160.
Khan, L. M. and Hanna, M. A., 1983, Expression of oil from oilseeds--A review: Journal of
Agricultural Engineering Research, v. 28, p. 495-503.
Knothe, G. and Steidley, K. R., 2005, Kinematic viscosity of biodiesel fuel components
and related compounds. Influence of compound structure and comparison to petrodiesel
fuel components: Fuel, v. 84, p. 1059-1065.
Reference list


54
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Krisnangkura, K., Yimsuwan, T., and Pairintra, R., 2006, An empirical approach in
predicting biodiesel viscosity at various temperatures: Fuel, v. 85, p. 107-113.
Lehada C.Shila. Director Vyahumu Trust. 7-2-2007.
Ref Type: Personal Communication
Mohibbe Azam, M., Waris, A., and Nahar, N. M., 2005, Prospects and potential of fatty
acid methyl esters of some non-traditional seed oils for use as biodiesel in India: Biomass
and Bioenergy, v. 29, p. 293-302.
Openshaw, K., 2000, A review of Jatropha curcas: an oil plant of unfulfilled promise:
Biomass and Bioenergy, v. 19, p. 1-19.
Paramathma, M., Reeja, S., Parthiban, K. T., and Malarvizhi, D. Biodiesel Conference
Towards Energy Independence - Focus on Jatropha:
Development of Inter-specific Hybrids in Jatropha. 10-6-2006. Forest college and
research institute, TNAU, Mettupalayam.
Ref Type: Conference Proceeding
Peter Chisawillo. Director Intermech. 7-2-2007.
Ref Type: Personal Communication
Pramanik, K., 2003, Properties and use of jatropha curcas oil and diesel fuel blends in
compression ignition engine: Renewable Energy, v. 28, p. 239-248.
Ramaswamy, R., Balasubramaniam, V. M., and Sastry, S. K., 2007, Thermal conductivity
of selected liquid foods at elevated pressures up to 700 MPa: Journal of food engineering,
v. 83, p. 444-451.
Romijn, H. and Biemond, I. Cost-Benefit Analysis of Projects. 2005. Eindhoven
University of Technology; Faculty of Innovation Sciences.
Ref Type: Report
Ruud van Eck. Diligent. 2006.
Ref Type: Personal Communication
Shahidi, F., 2005, Bailey's Industrial Oil & Fat Products volume 5: "Edible oil & fat
products: Processing Technologies": New Jersey, John Wiley & Sons, Inc..
Singh, J. and Bargale, P. C., 1990, Mechanical expression of oil from linseed (linum
usitatissimum L): Journal of oilseeds, v. 7, p. 106-110.
Singh, J. and Bargale, P. C., 2000, Development of a small capacity double stage
compression screw press for oil expression: Journal of food engineering, v. 43, p. 75-82.
Sirisomboom, P. et al. Physical and mechanical properties of Jatropha curcas L. fruits,
nuts and kernels. Biosystems Engineering . 2-22-0007.
Ref Type: In Press
Sukumaran, C. R. and Singh, B. P. N., 1989, Compression of a bed of rapeseeds: The oil-
point: Journal of Agricultural Engineering Research, v. 42, p. 77-84.
Vadke, V. S. and Sosulski, F. W., 1988, Mechanics of oil expression from Canola: Journal
of the American Oil Chemists' Society, v. 65, p. 1169-1176.
Vadke, V. S., Sosulski, F. W., and Shook, C. A., 1988, Mathematical simulation of an
oilseed press: Journal of the American Oil Chemists' Society, v. 65, p. 1610-1616.
Reference list


55
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries
Venter, M. J., Willems, P., Kuipers, N. J. M., and Haan, A. B., 2006, Gas assisted
mechanical expression of cocoa butter from cocoa nibs and edible oils from oilseeds: The
Journal of Supercritical Fluids, v. 37, p. 350-358.
Vyas, D. K. and Singh, R. N., 2007, Feasibility study of Jatropha seed husk as an open
core gasifier feedstock: Renewable Energy, v. 32, p. 512-517.
Ward, J. A., 1976, Processing High Oil Content Seeds in Notinuous Screw Presses:
Journal of the American Oil Chemists' Society, p. 261-264.
Willems, P., 2007, Gas Assisted Mechanical Expression Of Oilseeds: Ph.D. Thesis,
University of Twente, the Netherlands. ISBN: 978-90-365-2502-2.
Zheng, Y. l., Wiesenborn, D. P., Tostenson, K., and Kangas, N., 2003, Screw pressing of
Whole and Dehulled Flaxseed for Organic Oil: Journal of the American Oil Chemists'
Society, v. 80, p. 1039-1045.
Zheng, Y. l., Wiesenborn, D. P., Tostenson, K., and Kangas, N., 2005, Energy analysis in
the screw pressing of whole and dehulled flaxseed: Journal of food engineering, v. 66, p.
193-202.


56
Appendix A: Properties of Jatropha seeds and oil

Different parts of the Jatropha plant can be used as either solid or liquid fuel. The energy
content and recovery percentage of several Jatropha products are shown in Table A-1.
Recovery percentage of jatropha oil is expressed with respect to the whole fruit weight.

Table A-7-1 Energy values of varies Jatropha products and (Openshaw, 2000, 1-19).
Fuel Ash content (wt% d.b.) Moisture content (wt% w.b.) Energy value (MJ/kg) Recovery percentage
wood 1 15 15.5 95-100
whole fruit 6 8 21.2 95-100
whole nut 4 5 25.5 67-70
coat 13 15 11.1 28-30
hull 5 10 17.2 23-24
kernel 3 3 29.8 44-46
wood charcoal 3 5 30 15-25
hull charcoal 15 5 26.3 15-25
plant oil < 0.1 0 40.7 11-18
press cake 4 3 25.1 29-35


The composition of different components of the seed/nut are

Table A-2 gives an indication of the composition of seed and press cake. Approximately
34 wt.% of the seed is hull and the other 66 wt.% consists of the seed kernel
(Openshaw, 2000, 1-19).

Table A-7-2 Composition of Jatropha seed and press cake (Gubitz, Mittelbach, and Trabi, 1999, 73-
82)
1
(Openshaw, 2000, 1-19)
2
(Akintayo, 2004, 307-310).
Parameter
seed
1
kernel shell press cake
crude protein (%) 24.6-27.4 22.2-27.2 4.3-4.5 56.4-63.8
lipid (%) 48.7-51.4 56.8-58.4 0.5-1.4 1.0-1.5
ash (%) 4.6-4.9 3.6-4.3 2.8-6.1 9.6-10.4
neutral detergent fibre (%) 3.5-3.8 83.9-89.4 8.1-9.1
acid detergent fibre (%) 10.2-11.2 2.4-3.0 74.6-78.3 5.7-7.0
acid detergent lignin (%) 0.0-0.2 45.1-47.5 0.1-0.4
Gross energy (MJ/kg)
25.5
2
30.5-31.1 19.3-19.5 18.0-18.3


Some of the physical and chemical characteristics of Jatropha oil are represented in Table
A-3. Striking is the much lower viscosity encountered by Akintayo compared to Agarwal
(17.1 vs 35.98 cST). The Iodine value given in indicates application in the manufacturing
of alkyd resin, shoe polish, varnish etc. The high saponification value denotes good
properties for soap and shampoo production.
Table A-3 Physico-chemical characteristics of Jatropha oil (Akintayo, 2004, 307-310).
Parameter Jatropha oil
Colour Light yellow
Free fatty acid (mg/g) 1.76±0.10
Acid value (mg KOH/g) 3.5±0.1
Saponification value (mg KOH/g) 198.85±1.40
Iodine value (mg iodine/g) 105.2±0.7
Mean molecular mass (g) 281.62
Unsaponifiable matter (%) 0.8±0.1
Refractive index (25 °C) 1.468
Specific gravity (25 °C) 0.919
Hydroxyl value 2.15 ±0.10
Acetyl value 12 2.16±0.10
Viscosity (30°C) cST or mm
2
/s
17.1



57
Appendix B: German quality standard for Pure
Plant Oil






58
Appendix C: Data on BT 50 bio press setup
Dimensions and other specifications of the BT50 biopress are shown in Table C-1.
Channel height and width are indicated by H and W in the right side of Figure C-1. Angle
φ is visible in the left image of Figure C-1.

Table C-1 BT50 specifications
Srew dimensions
Screw diameter 48.6mm
Screw length 160mm (groove), 173 total length in contact with seed
Channel height H (constant) 11.5mm
Channel width W 11mm
Number of parallel groves 7 (6 full windings)
Flight thickness 5.75mm top, 11.6mm bottum
Angle φ 8 degrees
Cage diameter 50mm
Cage length 195mm
Other specifications
RPM 0-70
Engine nominal power 1.1kW





Figure C-1 Visualization of screw dimensions (Eggers, Broeck, and Stein, 2006, 494-499)

Pictures of the pressure sensor are visible in Figure C-2.




Figure C-2 Kistler 6152 mold cavity pressure sensor, amplifier and mounting method (Kistler
6152 user manual)




59
Appendix D: Maximum power output and
thermocouple calibration
The peak in Figure D-1 shows the Altivar 31 power output signal in Watts at the moment
the BT50 is jammed. The power indicated is around 2200-2300 W, which is twice the
nominal power of the electric motor as predicted by the Altivar 31 producer. This proves
that values from the variable speed drive are reliable.
0 20 40 60 80 100 120 140
0
500
1000
1500
2000
2500
P
o
w
e
r

o
u
t
p
u
t

(
W
)
Time (sec)
Power
0
20
40
60
80
100
120
140
T
e
m
p
e
r
a
t
u
r
e

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
Temperature

Figure D-1 Maximum power peak at press jamming

The type J thermocouples were calibrated using a thermostatic bath and a liquid mercury
thermometer. Results are shown in Figure D-2. Accuracy at low temperatures was within
1%, while for higher temperatures 2.5% was found.

40 50 60 70 80 90 100
40
50
60
70
80
90
100
110
T
h
e
r
m
o
c
o
u
p
l
e

v
a
l
u
e
s

(
d
e
g
r
e
e
s

C
e
l
s
i
u
s
)
Thermostatic bath temperature (degrees Celsius)
Thermocouple 1
Thermocouple 2
Thermocouple 3
Thermocouple 4
Thermometer

Figure D-2 Thermocouple calibration graph


60
Appendix E: Measurement equipment in Tanzania
The figures below give the reader an impression of the measurement equipment used for
soxhlet extraction and the determination of moisture content during experiments in
Tanzania. Equipment used for the determination of moisture content is shown in Figure
E-1.


Temperature profile heating
0
20
40
60
80
100
120
0 20 40 60 80 100 120
t (min)
T

(
c
i
n
t
e
g
r
a
d
e
)


Figure E-1 equipment used for the determination of moisture content from left to right: oven,
oven drying profile, balance (0.01 g accuracy).

Equipment used for soxhlet extraction is shown in Figure E-2. Figure E-3 shows that the
accuracy of the determination of foot levels was mainly due to the type of storage
container available.



Figure E-2 equipment used for soxhlet extraction from left to right: mortar and pestle to grind
samples before extraction, soxhlet setup, filter paper instead of extraction thimble.



61



Figure E-3 Left: oil containers. Right: determination of oil volumes and foot levels performed by
Diligent co-workers.



62
Appendix F: Sayari expeller and Intermech
workshop
Designs of the Sayari screw and press cage are shown more clearly in Figure F-1. One
can clearly see that transport sections and pressure cones are alternately positioned
along the worm shaft. Rotational speed of the Sayari expeller was 55RPM powered by a
standard 5.5kW electric motor (MA132SA4, MarelliMotori, Vicenza, Italy).



Figure F-1 Detailed views of the Sayari expeller. Left: the press cage made from separate lining
bars. Right top: worm positioned in the bottom part of the cage. Right bottom: separate worm
with clear distinction between cones and transport screw. The wheel on the left is for manual
adjustment of the outlet restriction.
Production of the press took place at the Intermech workshop (Morogoro, Tanzania). This
workshop produces several machines for agricultural applications. The assembly and final
product look as shown in Figure G-2.


Figure F-2 Assembly of Sayari expeller at Intermech workshop and final product.





63
Appendix G: Overview of variables included in this
research
The table below shows all the variations included in the present study. Limitations were
encountered by trial and error and were determined by the operational restrictions of
both presses. The table shows that the Sayari expeller could only be operated at a single
speed. Less hull fractions were tested in Tanzania because sample sizes were to big (10-
20 kg) for manual de-hulling. A wider range of moisture contents was tested at
Eindhoven University as better equipment was available to condition the seeds. The
temperature ranges for pre-heating the seeds were restricted to the point where no
visible alteration of the structure of the kernel was observed. The moisture content of
these pre-heated samples was not determined.


BT50 (Netherlands) Sayari (Tanzania)
Speed Speed
28RPM 56RPM
42RPM
49RPM
56RPM
70RPM
Restriction Restriction
7mm 4.5mm
9mm 3.7mm
12mm 2.7mm
1.3mm
0.6mm
Hull fraction Hull fraction
66% 60%
80% 100%
90%
100%
Moisture content Moisture content
2.14% 6.6%
5.40% 7.6%
6.30% 8.0%
6.70% 8.7%
7.30% 9.3%
8.50%
9.00%
13.30%
In addition In addition
Crushed Crushed
Torrefaction (100, 120, 140, 160, 180, 200 °C) Preheating in water 80°C
Preheating in water (70 & 80 °C)


64
Appendix H: Calculation of oil recovery from
soxhlet extraction
The following equations are required to calculate how much oil is recovered from the
seeds after pressing. The residual oil content in the press cake is compared to the oil
content in the seeds before pressing.

S O
O
W
+
=
0
0
0
,
S O
O
W
s
s
s
+
= ,
0
1
O
O
Y
s
− =

with:
Oo amount of oil in Jatropha seed sample (g)
Os amount of oil in Jatropha press cake sample (g)
S amount of solid material present in the sample
Wo oil content in original (g/g)
Ws oil content in sample (g/g)
Y oil recovery (or oil yield) (g/g)

Os and Oo are determined through soxhlet extraction. (Oo+S) the sample weight before
soxhlet extraction. The difficulty is that Os en Oo were determined from samples with
differing weight and can therefore not directly be inserted to the formula for Y.

The starting formulas can be rewritten into the following expression:

(
(
(
(
¸
(

¸

⋅ |
¹
|

\
|

⋅ |
¹
|

\
|

− =
S
Wo
Wo
S
Ws
Ws
Y
1
1
1

The S can be removed from the formula because the amount of solid material before and
after pressing is assumed equal. The S can be crossed out as the amount of oil is
compared per gram of solid material.




65
Appendix I: Initial measurement data
The graphs represented below give an indication of the accuracy of the output signals for
temperature, power and pressure. The temperature graph in Figure I-1 shows only slight
fluctuations in the range of ±5%.
-2 0 2 4 6 8 10 12 14
45
50
55
60
65
70
75
80
85
90
95
P
r
e
s
s
u
r
e

(
b
a
r
)
Time (min)
normal (6.7% moisture)
6.3% moisture
7.2% moisture
8.5% moisture
9.0% moisture
13.3% moisture

Figure I-1 Temperature profiles for different moisture contents at standard conditions for BT50
press (9mm, 70RPM, 100% hull)

The power data closely resembles pressure data as can be seen when comparing Figure
I-2 and Figure I-3. The output signals of power and pressure are more turbulent and can
vary about ±10%. Average values of graphs similar to the ones in this appendix were
used to construct the results in chapter 5.
-2 0 2 4 6 8 10 12 14 16
0
100
200
300
400
P
o
w
e
r

(
W
a
t
t
)
Time (min)
normal (6.7% moisture)
6.3% moisture
7.2% moisture
8.5% moisture
9.0% moisture
13.3% moisture

Figure I-2 Power requirement at different moisture contents at standard conditions for BT50 press
(9mm, 70RPM, 100% hull)


66
-2 0 2 4 6 8 10 12 14
20
30
40
50
60
70
80
90
100
110
120
130
140
P
r
e
s
s
u
r
e

(
b
a
r
)
Time (min)
normal (6.7% moisture)
6.3% moisture
7.2% moisture
8.5% moisture
9.0% moisture
13.3% moisture

Figure I-3 Smoothed pressure curves for different moisture contents at standard conditions for
BT50 press (9mm, 70RPM, 100% hull)



67
Appendix J: Calculation of energy requirement
The energy equation for the BT50 press provides insight in the amount of energy that is
needed to separate the oil form the solids. Comparing the required energy input to the
energy content of the oil results in an indication of the energy efficiency of the
mechanical extraction process. The calculations should be interpreted as a rough
calculation due to the use of natural materials like the Jatropha seeds. Properties like oil
yield and specific heat value can easily vary per batch. The steady state energy balance
is as follows:

rad cond conv o p o c c p c s s p s
Q Q Q T C M T C M W T C M + + + + = +
0 , , ,


where M is the mass flow in kg/s, C
p
is the specific heat value in J/kgK, T temperature in
°C and W the mechanical energy input to the screw in J/s. Subscripts ‘s’, ‘c’ and ‘o’ refer
to seeds, press cake and oil respectively. The Q terms indicate heat losses through
convection, conduction and radiation in J/s. In addition the mass balance equation is:

c s
M M M + =
0


Solving the energy equation results in an approximation for W. This value is compared to
the value measured by the volt output of the Altivar 31 (variable speed drive) that
regulates the power supplied to the BT50’s electric motor. For proper analysis of the
energy balance of pressing Jatropha seeds in the BT50 screw press the specific heat
values have to be properly estimated. One method to approximate C
p
-values is by using
the following expression (Zheng and others, 2005, 193-202):

t a f p c p
m m m m m C 187 . 4 837 . 0 675 . 1 549 . 1 424 . 1 + + + =

where m is the mass fraction on a wet basis and the subscripts indicate the following: ‘c’,
carbohydrate; ‘p’, protein; ‘f’, fat; ‘a’, ash; and ‘t’ water. M, T and W are easily
measured. The calculation of specific heats of seeds and cake is based on the
composition shown in Table J-1. The value for Jatropha oil was based on values for
Canola (Ramaswamy, Balasubramaniam, and Sastry, 2007, 444-451).

Table J-1 Composition of Jatropha seed and press cake used to calculate Cp
Component Jatropha seed cake
Crude fat 0.47 0.6
Crude protein 0.25 0.01
Crude fibre 0.10 0.1
Moisture 0.06 0.085
Ash 0.05 0.065
Carbohydrate (by difference) 0.08 0.14


The amount of energy needed per amount of seeds is expressed by the term ‘specific
mechanical energy (SME), which is commonly used for extrusion processes. SME denotes
the amount of energy that remains within the products and is calculated by rearranging
the energy equation (Zheng and others, 2005, 193-202):

s s s p s o o p o c c p c s rad cond conv
M T C M T C M T C M M Q Q Q W SME / ) , ( / ) (
, , ,
− + = − − − =

Because of the low temperatures, radiation heat losses are negligible compared to
convective and conductive heat losses and are therefore not taken into account.



68
The conductive heat losses can be estimated using Fourier’s law for steady state one-
dimensional heat conduction (Zheng and others, 2005, 193-202):

dl
dT
A k A k Q
s s b b d
) ( + − =
with:
Q
d
= conductive heat loss
k
b
= k
s
= thermal conductivity of screw and barrel which are assumed to be made of
stainless steel
A
b
and A
s
= barrel and screw outer surface area
dT/dl = heat loss per unit of length

Convective heat losses are approximated by the following equation also taking into
account the variation in outside diameter along the press (Zheng and others, 2005, 193-
202):
25 . 0
32 . 1 |
¹
|

\
| −
=
D
T T
h
r hs

with:
h = convective heat loss
T
hs
= hot surface temperature
T
r
= reference temperature /room temperature
D = Outer diameter of hot surface

The actual values used for estimating required energy input for screw pressing are shown
in Table J-2. The calculation done using previously estimated Jatropha values show a
power requirement of 170 kJ/kg seed for normal seeds. For comparisons sake
calculations using known Cp-values of cottonseed result in 97kJ/kg as can be seen in
Table J-2. The value measured by the Altivar 31 is 112kJ/kg for normal seeds and
303kJ/kg for seeds with 2.14% moisture. The model of Zheng et al does not approach
the values measured during testing. This might be due to inaccurate Cp-values or
limitations of the model. No effort was therefore made to calculate power requirements
for the Sayari expeller.


69
Table J-2 Input values and results for the calculation of power requirement for Jatropha pressing in
the BT 50 press
normal seeds dry seeds 2.14% moidture
M
cake
(kg/s) 0.67 0.67
M
oil
(kg/s) 0.33 0.33
M
seed
(kg/s) 1 1
Cp
jatrophacake
(J/kgK) 1630 1630
Cp
jatrophaoil
(J/kgK) 1910 1910
Cp
jatrophaseed
(J/kgK) 1556 1556
Cp
cottoncake
(J/kgK) 1494 1494
Cp
cottonoil
(J/kgK) 1787 1787
Cp
cottonseed
(J/kgK) 2176 2176
T
cake
(K) 371.8 394.9
T
oil
(K) 358.5 369.9
T
seed
(K) 298.0 298.0
SME for Jatropha estimated Cp values (J/kg seed) 168337 200806
SME for cottonseed Cp values (J/kg seed) 94882 124771
Conductive heat losses
k (w/mC) 16.3 16.3
A barrel (m2) 0.0093 0.0093
A screw (m2) 0.0019 0.0019
DT/dl 125 343
Conductive losses (W) 23 63
Conductive losses (J/kg seed) 1173 3218
Convective heat losses
T
hotsurface
(degrees Celsius) 98 121.9
T
barrel
(degrees Celsius) 85 96.9
T
reference
(degrees Celsius) 25 25
h
presshead
7.25 7.79
h
barrel
7.28 7.61
L
barrel
0.03 0.03
D
barrel
0.065 0.065
L
presshead
0.065 0.065
D
presshead
0.08 0.08
Convective heat loss (W) 11 16
Convective heat loss (J/kg seed) 582 806
Power requirement for Jatropha estimated Cp (kJ/kg seed) 170 205
Power requirement for cottonseed Cp (kJ/kg seed) 97 129



70
Appendix K: Foot levels in oil from the Sayari
expeller and BT50 press
Foot levels of the oil produced by the Sayari expeller are shown in Table K-1. For the
BT50 values are provided in Table K-2. Foot content in both tables proved too high for
normal filtering techniques applied by Liquid Filtration Consultants (LFC Lochum BV,
Lochem, The Netherlands) and suggested by Ferchau (2000). Foot contents for oil from
the BT50 seem to be somewhat lower than for the Sayari judging from both tables. Some
of the BT50 samples showed an additional ‘slimy’ component located in between oil and
sediments. The composition of this ‘slimy’ component was not studies into further detail
as filtering seemed to remove most of it from the oil.

Table K-1 Foot levels/sedimentation for oil from the Sayari expeller
Sample description oil (l) foot (l) total (l) foot content colour
6.6% moisture press cake 0.87 0.53 1.4 37.9% normal
6.6% moisture seeds 1 0.7 1.7 41.2% light
total 6.6% moistre 1.87 1.23 3.1 39.7%
6.7% moisture press cake 1.47 0.44 1.91 23.0% light
6.7% moisture seeds 1.14 0.63 1.77 35.6% light
total 6.7% moisture 2.61 1.07 3.68 29.1%
8% moisture cake 0.75 0.16 0.91 17.6% normal
8% moisture seeds 1.13 0.69 1.82 37.9% normal
total 8% moisture 1.88 0.85 2.73 31.1%
8.8% moisture seeds 0.52 0.58 1.1 52.7% light
8.8% moisture press cake 0.57 0.43 1 43.0% light
total 8.8% moisture 1.09 1.01 2.1 48.1%
9.3% moisture seeds 0.42 0.52 0.94 55.3% light
9.3% moisture press cake 0.37 0.42 0.79 53.2% light
total 9.3% moisture 0.79 0.94 1.73 54.3%
moist 2 seeds 1.47 1.13 2.6 43.5% light
boiled seeds 2.33 1.32 3.65 36.2% normal
normal 0 turns 1.875 0.85 2.725 32.3% light
normal 1 turn 1.5 1 2.5 40.0% light
normal 2 turn 1.7 1.33 3.03 43.9% light
normal 3 turn 1.68 2 3.68 54.3% light
cracked seeds 0.76 0.55 1.31 42.0% normal
cracked cake 0.77 0.4 1.17 34.2% normal
total cracked 1.53 0.95 2.48 38.3%
dehulled (60%) seeds 0.49 0.57 1.06 53.8% light
dehulled (60%) cake 0.4 0.34 0.74 45.9% light
total dehulled 0.89 0.91 1.8 50.6%




71
Table K-2 Foot levels for oil from BT50
Sample description total (cm) oil (cm) foots (cm) slime' (cm) foot content slime' content total impurities
normal 28 RPM 9.5 6.8 2.7 3.8 28.4% 11.6% 40.0%
normal 42 RPM 9.7 6.7 3 4.2 30.9% 12.4% 43.3%
normal 49 RPM 10.2 6.8 3.4 4.4 33.3% 9.8% 43.1%
normal 56 RPM 9.6 6 3.6 4.4 37.5% 8.3% 45.8%
normal 70 RPM 10.1 5.7 4.4 4.7 43.6% 3.0% 46.5%
100% hull 10.2 6.8 3.4 33.3%
90% hull 5.1 3.4 1.7 33.3%
80% hull - -
66% hull 5.1 2.5 2.6 51.0%
0
13.3% moisture 6.7 2.7 4 59.7%
9% moisture - -
8.5% moisture 6.3 3.9 2.4 38.1%
7.3% moisture 8.1 5.5 2.6 32.1%
5.4% moisture 9.8 6.5 3.3 33.7%
100°C preheated 6.8 4.3 2.5 36.8%
120°C preheated 6.9 4.4 2.5 36.2%
140°C preheated 7.3 4.6 2.7 37.0%
160°C preheated 7.2 4.5 2.7 37.5%
180°C preheated 7.2 4.7 2.5 34.7%



72
Appendix L: Reinartz & Keller presses and
industrial oil recovery percentages

Machinenfabrik Reinartz
Neuss / Germany
Phone: +49 2131 9761-0
Fax: +49 2131 9761-12

The AP03, shown in Figure L-1, is particularly suited for soft oilseeds like rapeseed,
linseed, sunflower and jojoba. Tests were done under standard settings. When the
Jatropha seeds were de-hulled before pressing the amount of sediments appeared to
increase. Specific values on the amount of sediment have to be determined by amafiilter
(amafiltergroup Alkmaar B.V., Alkmaar, The Netherlands).



Figure L-1 Technical details of the AP08 expeller

KEK Keller
Remscheid / Germany
Phone: +49-2191-8410-24
Fax: +49-2191-8628

Tests with Jatropha seeds were executed to find out whether high quality expellers would
yield more oil than their Tanzanian counterparts. Details on the press used for these tests
are shown in Figure L-2. Two separate tests were carried out with Jatropha seeds from
Tanzania. The first test was under standard press settings for rapeseed and the second
was optimized for Jatropha by inserting two larger pressure cones and changing the feed
section for more transport capacity.


73

Figure L-2 Technical details of KEK P0101 expeller

Some detailed pictures of some of the Keller press components are shown in Figure L-3
to underline the quality difference with the Sayari expeller shown in Appendix F.



Figure L-3 Detailed view of some of the KEK P0101 components. From left to right: press cage
made from separate bars, worm with cones for pressure build-up, adjuster for outlet restriction

Table L-1 gives an impression on the residual oil content of industrial oil expellers for
various oilseeds. The residual oil contents are comparable to those achieved for Jatropha
in BT50, Sayar, AP08 and P0101 presses. This finding indicates that Jatropha processing
should be profitable.


74
Table L-1 Capacities and residual oil contents for three press types applied in industrial oil mills




75
Appendix M: CBA results for different process
setup
Decentralized processing
case 1: 5 x KEK P0101 (70kg/hr) 8 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$537,977 -$537,977 1.000 -$537,977
1 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.925 $22,655
2 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.856 $20,962
3 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.792 $19,395
4 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.733 $17,945
5 $364,076 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $388,561 0.678 $263,471
NPV -$193,550
IRR -2%
case 2: 5 x KEK P0101 (70kg/hr) 24 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$567,977 -$567,977 1.000 -$567,977
1 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.925 $197,227
2 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.856 $182,482
3 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.792 $168,840
4 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.733 $156,218
5 $379,076 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $592,238 0.678 $401,578
NPV $538,367
IRR 34%
case 3: 5 x KEK P0101 (70kg/hr) 8 hours operation per day with sedimentation (oil recovery is 0.2 instead of 0.3)
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$282,003 -$282,003 1.000 -$282,003
1 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.925 -$13,442
2 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.856 -$12,437
3 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.792 -$11,507
4 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.733 -$10,647
5 $189,548 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 $175,020 0.678 $118,675
NPV -$211,360
IRR -14%
case 4: 5 x KEK P0101 (70kg/hr) 24 hours operation per day with sedimentation (oil recovery is 0.2 instead of 0.3)
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$312,003 -$312,003 1.000 -$312,003
1 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.925 $46,749
2 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.856 $43,254
3 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.792 $40,020
4 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.733 $37,028
5 $204,548 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $255,073 0.678 $172,957
NPV $28,005
IRR 11%
case 5: 5 x Sayari (70kg/hr) 8 hours operation per day with cyclone filter and cheaper storage
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$299,575 -$299,575 1.000 -$299,575
1 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.925 $27,789
2 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.856 $25,711
3 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.792 $23,789
4 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.733 $22,011
5 $182,029 -$62,271 -$4,071 -$47,205 -$800 $144,382 $212,063 0.678 $143,793
NPV -$56,481
IRR 3%
case 6: 5 x Sayari (70kg/hr) 24 hours operation per day with cyclone filter and cheaper storage
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$329,575 -$329,575 1.000 -$329,575
1 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.925 $198,357
2 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.856 $183,528
3 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.792 $169,807
4 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.733 $157,113
5 $197,029 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $411,413 0.678 $278,966
NPV $658,196
IRR 63%






76

case 7: 5 x Sayari (70kg/hr) 8 hours operation per day with cloth filter and cheaper storage
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$51,850 -$51,850 1.000 -$51,850
1 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.925 -$8,308
2 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.856 -$7,687
3 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.792 -$7,113
4 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.733 -$6,581
5 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.678 -$6,089
NPV -$87,628
IRR -
case 8: 5 x Sayari (70kg/hr) 24 hours operation per day with cloth filter and cheaper storage
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$81,800 -$81,800 1.000 -$81,800
1 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.925 $46,749
2 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.856 $43,254
3 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.792 $40,020
4 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.733 $37,028
5 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.678 $34,260
NPV $119,510
IRR 55%

Centralized processing
case 9: 1 x Reinartz 14/30 (350kg/hr) 8 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$250,760 -$250,760 1.000 -$250,760
1 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.925 $15,945
2 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.856 $14,753
3 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.792 $13,650
4 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.733 $12,630
5 $165,518 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $182,751 0.678 $123,918
NPV -$69,865
IRR 0%
case 10: 1 x Reinartz AP 14/30 (350kg/hr) 24 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$310,760 -$310,760 1.000 -$310,760
1 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.925 $181,915
2 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.856 $168,315
3 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.792 $155,732
4 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.733 $144,090
5 $195,518 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $392,132 0.678 $265,892
NPV $605,185
IRR 61%
case 11: 1 x Reinartz AP15/45 (750kg/hr) 8 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$361,686 -$361,686 1.000 -$361,686
1 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.925 $84,491
2 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.856 $78,175
3 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.792 $72,330
4 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.733 $66,923
5 $211,240 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $302,558 0.678 $205,155
NPV $145,387
IRR 20%
case 12: 1 x Reinartz AP15/45 (750kg/hr) 24 hours operation per day with cyclone filter
Year Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow
0 -$481,686 -$481,686 1.000 -$481,686
1 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.925 $447,203
2 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.856 $413,770
3 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.792 $382,837
4 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $483,337 0.733 $354,216
5 $271,240 -$523,079 -$61,071 -$153,877 -$10,124 $1,231,487 $754,577 0.678 $511,654
NPV $1,627,994
IRR 99%



77
Appendix N: Calculation of the Engaruka case
Poor yield 100 acres
Year Equipment Running costs Revenues Total Discount factor Discounted cash flow
0 -$4,795 -$4,795 1.000 -$4,795.06
1 -$7,314 $13,095 $5,781 0.925 $5,348.71
2 -$7,314 $13,095 $5,781 0.856 $4,948.73
3 -$7,314 $13,095 $5,781 0.792 $4,578.67
4 -$7,314 $13,095 $5,781 0.733 $4,236.28
NPV $14,317.33
IRR 115%
Standard situation 100 acres
Year Equipment Running costs Revenues Total Discount factor Discounted cash flow
0 $0 1.000 $0.00
1 -$2,450 $6,300 $3,850 0.925 $3,562.10
2 -$2,450 $6,300 $3,850 0.856 $3,295.73
3 -$2,450 $6,300 $3,850 0.792 $3,049.28
4 -$2,450 $6,300 $3,850 0.733 $2,821.25
NPV $12,728.36
Poor yields 50 acres
Year Equipment Running costs Revenues Total Discount factor Discounted cash flow
0 -$4,795 -$4,795 1.000 -$4,795.06
1 -$4,002 $6,344 $2,343 0.925 $2,167.51
2 -$4,002 $6,344 $2,343 0.856 $2,005.42
3 -$4,002 $6,344 $2,343 0.792 $1,855.46
4 -$4,002 $6,344 $2,343 0.733 $1,716.71
NPV $2,950.03
IRR 33%
Optimistic yields 100 acres
Year Equipment Running costs Revenues Total Discount factor Discounted cash flow
0 -$4,795 -$4,795 1.000 -$4,795.06
1 -$23,132 $40,097 $16,965 0.925 $15,696.24
2 -$23,132 $40,097 $16,965 0.856 $14,522.49
3 -$23,132 $40,097 $16,965 0.792 $13,436.51
4 -$23,132 $40,097 $16,965 0.733 $12,431.73
NPV $51,291.91
IRR 353%

Standard situation 100 acres optimistic yield
Year Equipment Running costs Revenues Total Discount factor Discounted cash flow
0 $0 1.000 $0.00
1 -$2,450 $18,900 $16,450 0.925 $15,219.88
2 -$2,450 $18,900 $16,450 0.856 $14,081.75
3 -$2,450 $18,900 $16,450 0.792 $13,028.73
4 -$2,450 $18,900 $16,450 0.733 $12,054.45
NPV $54,384.80


78
Appendix O: Villages for case study
Case studies
Rural areas could potentially benefit from the products coming from the Jatropha plant.
However most of these products, except the seeds themselves, require processing
equipment for production. The example that leaps to the eye is oil production. In any
case a machine to press the oil from the seeds in necessary. Storage tanks and filters are
also required. Before the oil is ready to put into an engine neutralizing and filtering
should be conducted, although in case of older Lister engines crude filtering is sufficient.
To a lesser extent these investments also hold for soap, fertilizer and briquette
production. One can imagine that most rural communities in Tanzania are unable to
invest in such machinery and therefore end up selling yet another primary product at
unrewarding prices. In an attempt to depict to what extend local people can exploit the
benefits associated with Jatropha without a strong dependence on European and North
American investors three different Jatropha cases are developed. The cases represent
three villages in Northern Tanzania with a different level of potential to use Jatropha.
Their classification is as follows:

Temi ya Simba
This small village of approximately 600 inhabitants close to Arusha has no electricity grid
connection and for cooking the inhabitants use mostly wood, which they collect outside
the village. Charcoal stoves or the more sophisticated plant oil stoves are not yet
introduced to this region. Two farmers are growing Jatropha and around 100 people grow
trees in hedges throughout the village and around agricultural land. During the first
harvest 170 kg of Jatropha seeds were collected for Diligent and 600kg was used by the
women group for soap production. According to Mr Rubanze the total Jatropha production
potential for this village is around 4-6 tons/year two years from now (in case of two
harvest seasons per year). Local applications for Jatropha oil are substitution of
conventional diesel for the five village maize mills and soap production by a local women
group supported by Kakute. It is interesting to find out if the people are interested in
electricity, charcoal or even soap production on a larger scale. Error! Reference source
not found. shows diagram clarifying the Jatropha applications and the flow of goods
between actors in Temi ya Simba. The colours classify the benefits of Jatropha to a
certain actor. Farmers are the only village members assumed to generate additional
income form Jatropha. The user experience more practical advantages like increases fuel
availability and healthier cooking methods due to decreased smoke production.


Figure 0-1 Jatropha utilization in Temi ya Simba
Maji ya Chai
In the region of Maji ya Chai many people are enthusiastic about the introduction of oil
fuelled Bosch-Siemens Protos plant oil stoves by the German NGO GTZ. One reason for
the introduction of the stoves is to get people to use renewable oils instead of fossil fuels
or wood. The use of cooking wood is one of the most important causes fro deforestation
in Tanzania. However villagers cannot afford to pay for the oil to fuel the stove as
Jatropha oil and sunflower oil are more expensive than both wood and kerosene. This
village serves as a case to check if growing Jatropha solely for cooking purposes is


79
feasible. As can be seen in the model in Error! Reference source not found. the
farmers gain additional income by processing the seeds into oil at farm level.


Figure 0-2 Jatropha utilization in Maji ya Chai

Engaruka
The village of Engaruka consists of two separate parts: northern and southern Engaruka.
Most of the 9000 inhabitants of this Maasai village make a living as farmers or keeping
feedstock. As is the case for Temi ya Simba there is no grid connection present in the
village. Therefore a Multi Functional Platform (MFP) was installed by a Tanzanian NGO
named TaTedo in November 2006 to combine seed pressing with electricity generation
and grain milling. An impression of the MFP is given in Figure 7-1. The MFP operates five
hours a day generating roughly 6 kWh. In total 20 shops, 3 houses and a campsite are
provided with energy from this platform. In addition to the grain mill connected to the
MFP seven more diesel powered grain mills are present in the village. Together with the
diesel driven generator these machines constitute the potential Jatropha consumption in
Engaruka. As Jatropha was already grown in Engaruka for fencing most seeds still stem
from hedges. Over 30 farmers are growing Jatropha on small plots between 0.5-3 acres
in size(2007). The first harvest in 2007 produced approximately 2400 kg Jatropha seeds.
The total land area available for Jatropha is around 100 acres. As can be seen in Figure
6-2 the model for Engaruka shows most economical activities as a result of Jatropha
processing. The most important expansions are the inclusion of electricity production and
a larger group of people generating additional income from Jatropha. A press cooperation
consisting of maybe 5-10 people could start production of Jatropha oil and grain millers
could also benefit from this cheaper locally produced fuel to power their engines. The
investment required to setup a press-cooperation probably exceed the amount of money
available to the local investors. Most of the villagers cannot get any loans from the bank
as farming is excluded from the definition of enterprises the bank uses when judging loan
requests. Therefore Savings and Credit Co-operatives (SACCOs) are included in this
system. The SACCO system is a mutual membership organization, which involves pooling
of voluntary savings from members in the form of shares. These savings or shares form
the basis for providing credit to members. The available credit is normally around three
times the level of savings or shares. The importance of this market segment is underlined
by a SACCOs market of nearly US$1 billion in Kenya. (Enterprising Solutions Global
Consulting, 2003)



80

Figure 7-1 Left hand picture: A TaTEDO test MFP near Dar es Salaam with from left to right a Sayari
screw press, Lister engine and a maize mill. Between the engine and the maize mill normally a
generator is fitted like the blue one shown in the right hand picture that was taken in Engaruka.



Figure 7-2 Jatropha utilization in Engaruka

Acknowledgements

Acknowledgements
This thesis was written to graduate for the Master Sustainable Energy Technology at Eindhoven University of Technology. Occasion to this study was a question from Diligent Energy Systems on how to optimize Jatropha oil production. The study consisted of practical research at the university and a two and a half month visit to Tanzania for both practical research and field studies. I would like to use this opportunity to thank the people in the Netherlands and Tanzania who made my graduation largely a pleasant experience. If it had not been for the enthusiasm of Kees Daey Ouwens I would never have started this research about Jatropha and the pressing process in particular. I want to thank him for helping me to shape the project from the start, his feedback during the process and providing me the opportunity to visit Tanzania. Furthermore I am very glad that André de Haan was willing to supervise my project with such enthusiasm, thereby enabling me to research the topic that appealed to me. As I depended on the help of many persons I want to express my thanks by mentioning them here: • Ies Biemond and Mark Prins, my second and third supervisor • Paul Willems, for helping me with the setup of my practical research and the extensive feedback on this thesis • Ruud van Eck, for adding practical relevance to my research and enabling me to visit Diligent Energy Systems in Tanzania. • Janske van Eijck, for receiving me at Diligent Tanzania and making me feel at home. • All employees of Diligent Tanzania LTD, for their warm welcome and their help to complete my research. • Supporting staff members of Combustion Technology at the faculty of mechanical engineering; especially Frank Seegers and Gerard van Hout who helped me with most practical arrangements. • Supporting staff members of System Process Engineering at the faculty of chemical engineering, for helping me with the setup of my analysis methods. • Niels Ansø, for providing me with spare parts for the screw press free of charge. • All the people that I forgot to mention above who contributed to this thesis. • The companies Reinartz and Keller, for providing me with the results of their Jatropha press tests. Finally I want to thank my girlfriend, my friends and my parents who gave me the support, thrust and opportunities to distract my mind in order to complete my graduation. Peter Beerens, Einhoven, July 2007.

i
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

i
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

Summary

Summary
People living in the poorest regions in the world, e.g. large parts of the African, Asian and Latin American continents, often lack access to energy sources in general. One approach to provide these people with energy to increase living standards is to enable them to produce energy from local resources. A promising local renewable energy source for people living in tropical regions is Jatropha Curcas L; a plant that grows oil yielding seeds that contain 31-38% oil. The seeds need to be processed in order to obtain the useful end product oil. Mechanical screw pressing is the most popular method in the world to separate oil from vegetable oilseeds on small to medium scale. Therefore the impact of several variables on oil recovery, oil quality and energy costs for screw pressing is reviewed in this report. The main question to be answered in this report is: • How can oil recovery, energy input and cost of Jatropha seed expression in screw presses be optimized while retaining a quality that complies with international fuel standards?

Additional consideration of the practical application of screw presses for the production of Jatropha oil in Tanzania led to answers to the following two additional questions. • • What is the optimal setup for the production of Jatropha oil in Tanzania? In what way can rural areas in which their use is anticipated benefit from Jatropha cultivation?

Two distinct low capacity screw presses were used; one located in the Netherland and the other one in Tanzania. Comparing the results of test on both presses reveals which of the two types is best suited to extract oil from Jatropha seeds. The following variables were in advance considered of main interest: • • • • Size of the restriction at the end of the press chamber Rotational speed of the screw press shaft Moisture content of the seeds Hull content of the seeds

The most important conclusions that can be drawn from this study are as follows: Moisture content has the strongest effect on oil recovery. Restriction size and rotational speed of the screw are other influential parameters. The highest oil recovery measured were 89% and 91% for the BT50 and Sayari press respectively. These values resulted after one hour cooking in water of 70°C. Oil recovery values for untreated seeds under standard circumstances were 79% and 87%. Important to note is that the Sayari expeller requires dual passing of the material compared to single passing for the BT50. Additional Jatropha press tests were conducted at two industrial German press producers in order to put the other results into perspective. The tests conducted during this research showed close resemblance to industrial practice. Taking into consideration all the test results optimal oil recovery is expected at a moisture content of 2-4% after cooking at 70°C, 100% hull content and the smallest restriction size and lowest speed possible for a certain press type. Based on purely financial results from the ‘cost benefit analysis’ large scale centralized processing is preferred over decentralized pressing. For similar production scale centralization allows an ‘internal rate of return’ of 61% compared to 30-40% in case of decentralization. However, reviewing both financial and non-financial aspects shifts preference to decentralized pressing for Tanzania and similar countries.

ii
Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries

100 acres of Jatropha could easily provide an additional $1. which is equivalent to 19 monthly salaries to be divided between the initiators.500 depending on seed yield.500 per annum. which is equivalent to approximately 100 annual minimum wages. For a village of 9000 inhabitants Jatropha cultivation has an added value between $12.750 and $54.Summary Jatropha has significant added value for rural communities in Tanzania. Oil processing on village scale without the intervention of commercial firms seems preferable over mere seed selling if coupled to electricity generation and maize milling. iii Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .

2 The two main screw press designs 3.2 Seed properties 2.2 Influence of RPM 5.4 Analysis 4.11 Conclusions 6 Possibilities for oil pressing in rural Tanzania 6.1 Data processing 5.3.2.1 Materials 4.1.3.2 Cylinder-hole press 3.1 Strainer press 3.1 Introduction i ii 1 1 1 1 2 2 3 3 4 5 5 6 7 7 8 9 11 11 14 14 14 15 16 16 16 18 18 18 18 19 20 21 21 21 21 22 23 24 25 26 27 28 30 32 34 35 35 36 36 36 37 39 39 iv Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .3.1.2 Problem 1.9 Oil quality parameters 5.2.1.1 Research question 1.2 Biodiesel 3 The screw press 3.2 Working Method 1.1 Experiments at Eindhoven University of Technology 4.8 The effects on foots 5.3 Different designs 3.4 Influence of moisture content 5.1 Independent variables 3.3.2.3.1.1.Table of content Table of content Acknowledgements Summary 1 Introduction 1.3 Parameters affecting the screw pressing process for oilseeds 3.2.1.1.2 Experimental setup 4.2 Experiments at Diligent Tanzania LTD 4.7 Influence of other pre-treatments 5.2.3 Jatropha applications 2.5 Accuracy and reproducibility of experimental results 5 Results and discussion 5.2.2 Goals 1.6 Influence of seed preheating 5.2 Experimental setup 4.10.2.10 Comparison of suitability for Jatropha processing between BT50 and Sayari 5.2.3 Experimental procedure 4.2 Dependent variables 4 Materials and method 4.4 Analysis 4.2.3 Influence of restriction size 5.2.3 Report structure 2 Jatropha and its applications 2.1.1 Occasion 1.1 Background 1.3 Research limitations 1.3 Experimental procedure 4.1.1 Working principles 3.1 Materials 4.5 Influence of hull fraction 5.5 Accuracy and reproducibility of experimental results 4.1 Plant description 2.1 PPO 2.2.1 Test conducted by industrial press manufacturers 5.1 Fuel 2.

Table of content 6.3 Sensitivity analysis 6.2 Cost Benefit Analysis: Theory 6.3 CBA results: comparison of centralized and decentralized production 6.2.3.5 Conclusions 7 Conclusions and recommendations Reference List Appendix A: Properties of Jatropha seeds and oil Appendix B: German quality standard for Pure Plant Oil Appendix C: Data on BT 50 bio press setup Appendix D: Maximum power output and thermocouple calibration Appendix E: Measurement equipment in Tanzania Appendix F: Sayari expeller and Intermech workshop Appendix G: Overview of variables included in this research Appendix H: Calculation of oil recovery from soxhlet extraction Appendix I: Initial measurement data Appendix J: Calculation of energy requirement Appendix K: Foot levels in oil from the Sayari expeller and BT50 press Appendix L: Reinartz & Keller presses and industrial oil recovery percentages Appendix M: CBA results for different process setup Appendix N: Calculation of the Engaruka case Appendix O: Villages for case study 39 40 41 41 41 43 44 45 45 46 48 50 52 56 57 58 59 60 62 63 64 65 67 70 72 75 77 78 v Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .3.4 Viability of community level processing: Engaruka case 6.2.2.3.2 Centralized pressing 6.3 Sensitivity analysis 6.2 Liquidity assessment 6.1 Decentralized pressing 6.4 Non financial consideration 6.3.1 Profitability assessment 6.

2000.1 Occasion 1. In addition changes in moisture content and physical. which could lead to long term improvement of well-being. To name a few countries: Belize. 130-134). 2000. 130-134).Chapter 1: Introduction 1 Introduction This chapter provides some background information important for this research. Both the oil and the de-oiled press cake obtained using screw presses are free of solvents and other chemicals as opposed to the more efficient solvent extraction method. temperature and rotational speed of the press worm. taking into consideration the possible application in rural Africa. By means of such an approach the reliance on help from foreign countries is limited. can easily be operated. Some examples are use for fencing. 1169-1176. 130-134. electricity. Ghana.D.Zheng and others. Reasons for its popularity in for example India are that the machines require low initial and operation investments. fertilizer and soap (Openshaw.1. 2004. a plant that grows oil yielding seeds. 193-202). large parts of the African. One approach to provide these people with energy to increase living standards is to enable them to produce energy from local resources. 2005.g. Finally the structure of this report is shortly described. 1. 1. 2000. 2000. 1-19). 261-264. There is however a 1 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .Henning. Mechanical screw pressing is the most popular method in the world to separate oil from vegetable oilseeds on small to medium scale.2 Problem Although wild varieties of Jatropha exist in many regions attempts to cultivate the plant are limited. 2003. A promising local renewable energy source for people living in tropical regions is Jatropha Curcas L. Once the oil is removed from the seeds. Asian and Latin American continents. The effect of research outcomes was that oil recovery rates increased from 50% to 80% for various oilseeds (Bargale and Singh. Research on mechanical screw presses for pressing oilseeds dates back to 1951 when V. Mali and Mozambique (Hartlieb Euler and David Gorriz. Ever since considerable efforts have been made to gain better understanding of the processes inside a screw press and improve the oil extraction efficiency (Bargale and Singh. 1976. Tanzania.Vadke and Sosulski.1 Background People living in the poorest regions in the world. Based on previous research various pre-treatments and process variables are considered for the present study.Ward. maintained and adopted by semi-skilled personnel (Bargale and Singh. Working method and research limitations are also shortly elaborated on. Nicaragua. 2007).1. Furthermore the research problem and goals are stated. Anderson Company patented the first expeller. gas. decreased deforestation for firewood collection and reduction of soil erosion. 10391045. the oil and its by-products can serve various purposes some of which are: liquid fuel. e. often lack access to energy sources in general. Zambia.Zheng and others. 1988. India and Latin America over the last few years. Furthermore it is possible to manufacture screw presses locally creating additional local employment. Cultivation experiments have been carried out in the previous decade and actual plantations have emerged in Africa. For cooking they often depend on gathered fuel wood and in case a connection to the electricity grid is present it is often unreliable. The earlier research was aimed at optimizing process variables like pressure. either mechanically or chemically. thermal and hydrothermal pretreatment of the seeds were considered in later research. What distinguishes Jatropa from many other bio fuel crops are the benefits it can offer to relatively small rural areas in Less Developed Countries.

In 2003 Diligent started to contact local farmers and create networks for the collection of Jatropha seeds. energy input and cost of Jatropha seed expression in screw presses be optimized while retaining a quality that complies with international fuel standards? In this question oil recovery is defined as the amount of oil removed from the seeds compared to the amount that was initially present in the seeds. By the end of 2006 their supply of seeds became sufficient to go on to the next step. A moment of reflection reveals that this challenge not only requires a technical solution. which is processing the seeds into high quality oil.Chapter 1: Introduction lot of uncertainty on the best way to process the seeds into oil. Previous studies showed that oil recovery rates can be raised from 73% to 80% of the initial oil content for rapeseed and peanut and from 60% to 65% for cotton seeds by improvement of the press settings and proper the economic viability of local oil production by screw pressing and should therefore also be examined for Jatropha. but still failed dramatically (Douthwaite. The problems Diligent encountered in determining the best suited processing method for Jatropha seeds in Tanzania gave occasion to this research. electricity (generator).1 Research question The most interesting application for the Jatropha oil is to use it as a fuel. It might therefore be worth the effort to go in search of an appropriate implementation approach for each specific project. For this research the north-eastern region of Tanzania was selected as a case study. One of the disadvantages of screw pressing is the low oil recovery when processing untreated seeds. Numerous examples exist of innovations that where technically sound. 2005).2 Goals 1.2. the production process can be chosen and analysed. taking into account the backward circumstances in less developed countries. Therefore these are also the aspects touched upon in the main research question: • How can oil recovery. lighting and cooking. 1. changes in press operation and changes in press design. This way the oil can be used for mobility (vehicles). 2002). Therefore a second and third question are included to identify and avoid some of the pitfalls correlated with the introduction of innovations in less developed countries: • • What is the optimal setup for the production of Jatropha oil in Tanzania? In what way can rural areas in which their use is anticipated benefit from Jatropha cultivation? 2 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . The three most important aspects to consider when analysing expeller operation are the amount of oil removed. However this research is extended beyond the determination of the optimal condition under which Jatropha seeds should be processed. but also demands investigation of local circumstances. Once the intended use of the product is known. The success of a new technology or production process is inseparably linked with the way it is introduced. In and around the cities of Arusha and Moshi attempts are made to set up a complete system to grow. There are three possible approaches to improve the properties mentioned in the research question namely pre-treatment of seeds before pressing. process and use Jatropha to its full potential. Diligent Energy Systems is a Dutch company operating in the field of bio fuel production in less developed countries and active in Tanzania and Colombia. the amount of energy required to achieve this and the quality of the oil with respect to the intended application. Even after multiple passes 5% to 10% of the oil remains in the press cake after pressing (Shahidi.

The main goal of the visit was to identify local circumstances and technological means to design a plan for local expression of Jatropha oil.3 Research limitations In this report the impact of several process variables on oil recovery. testing all fuel quality parameters is outside the scope of this graduation project. Based on literature the most important variables influencing the process will be (Bargale and Singh. Field studies were conducted from Jan-Mar 2007 consisting of visits to farmers who grow Jatropha. At the moment of writing there is little activity involved in organized pressing of the seeds. Broeck.Eggers. In addition costs and maintenance required for operation are estimated. running Jatropha presses. The results are obtained by practical tests on two different standard types of screw presses. oil quality and energy costs are reviewed. The extent to which oil is separated from solid material is determined by analysing and comparing the oil content of both input (seeds) and output material (press cake). and Stein. 11691176. The practical research is restricted to variation of pre-treatments and press operation. Comparing the results of test on both presses reveals which of the two types is best suited to extract oil from Jatropha seeds. 1. 1988. 3 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .2 Working Method Two distinct low capacity screw presses were used. involved Non government Organisations (NGOs). 2005).2. universities actively researching Jatropha and other interesting opportunities that presented itself.2. The only method of oil extraction examined in the report is mechanical extraction. a local producer of screw presses. one located in the Netherland and the other one in Tanzania. 2003. For the purpose of pressing oilseeds on a small scale in less developed countries screw presses seem to be the better option despite their lower yield. 1039-1045): • • • • Size of the restriction at the end of the press chamber Rotational speed of the screw press shaft Moisture content of the seeds Hull content of the seeds Because Jatropha oil is intended to be used in diesel engines it is important to compare the quality of the oil according to international standard for pure plant oils (PPO) and biodiesel. 2000. In spite of its higher oil yields solvent extraction has some drawbacks especially for small scale applications.Zheng and others. Solvent extraction can achieve a reduction in the oil content of the press cake to less than 1% (Shahidi. As current methods of solvent extraction are not applicable in the region of interest they are not treated in further detail in this study. Changes in screw design in order to improve oil yield were not included in this study. Pre-treatments for size reduction and cell rupturing are also required prior to solvent extraction adding extra costs. Both screw press settings and properties of the Jatropha seeds are varied to get a better understanding of the expression process and to find the optimal process conditions. During a visit to Tanzania the test results gathered at Eindhoven University of Technology were compared to a different type of screw press operating at Diligent’s workshop in Arusha. 494-499.Chapter 1: Introduction 1. In addition the commercially used solvent hexane can become highly explosive when mixed with air. As this involves many different and time consuming chemical analyses. namely cylinder-whole press and strainer press. 2005). Indispensable tests on for example acidity and phosphorous content were contracted out to external research institutes. Tanzania. 130134. Solvent extraction facilities are typically constructed to process several thousands of tons of seed per day and cost in the order of tens of million dollars (Shahidi. 2006.Vadke and Sosulski. The collected data was used to perform a ‘Cost Benefit Analysis’ (CBA) on optimal Jatropha processing scale in Tanzania.

Based on the identification of local circumstances and a Cost Benefit Analyses (CBA). The research results are explained and discussed in chapter 5. The examination of socio-technological aspect in addition to the technical aspects is considered an added value to this report.Chapter 1: Introduction Further treatment of the oil that comes out of the press might be required before it can be used to properly run a diesel engine. Finally conclusions and recommendations are given in chapter 7. Although these steps are touched upon in the report no tests were carried out on the effect of different treatments of the oil on the performance or emission of a diesel engine. In this third chapter the variables affecting the pressing process are also discussed. In chapter 3 the working principles of different types of screw presses are explained. ideas on how to setup Jatropha pressing in Tanzania is discussed. Presumably the performed CBA will be less thorough than would be the case for a graduation project fully dedicated to this subject. Chapter 6 provides an overview of the findings in Tanzania. Chapter 4 focuses on the research method of the experiments that were carried out on the screw press and the chemical analyses used to analyse the press cake. 4 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . 1.3 Report structure The Jatropha plant and the applications of the oil and its by-products are elaborated on in more detail in chapter 2.

The plant belongs to the family Euphorbiaceae and is indigenous to Latin America (Akintayo.2 some interesting seed properties with regards to screw pressing are mentioned. Jatopha is a bush or small tree that is able to survive on marginal lands and can get up to 6 or 8 meters high. will for simplicities sake from now on be referred to as Jatropha. the most common of which is “physic nut”.h) and seeds (i) Right: close-up of Jatropha fruits near Arusha. fruits (g. 2006). Being drought tolerant the plant still sheds its leafs during dry periods. usually they refer to the species Jatropha Curcas L.Chapter 2: Jatropha and its applications 2 Jatropha and its applications This chapter serves as both an introduction to Jatropha for the layman and a detailed description of characteristics of Jatropha seeds and oil related to screw pressing. In paragraph 2. 1-19). 2002. and Srikanth. 161-164). Seeds from the Jatropha tree are known by many different names. 2000. 2000. which is one of the 170 known species of this plant (2006. In Tanzania it is known as “Mbono” or “Makaen” and in the Netherlands as “purgeernoot”. Jatropha is a wild plant. Ideal would be 600mm in moderate climates and 1200mm in hot climates (Biswas. Jayabalan. of particular interest for this study. In addition to water a plant needs 5 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . 1996). 161-164). Under favourable conditions it can grow to a thick bushy fence of approximately one meter high in 6-9 months (Augustus. 2004. in other words it has not been cultivated through variety research. 2. This perennial Figure 2-1 Left: Representation of the Jatropha plant (a.c. 307-310) and naturalized throughout tropical and subtropical parts of Asia and Africa where it is mostly used for hedging (Augustus. Kaushik.Paramathma and others. and Seiler.b. and Seiler. Paragraph 2. which means that it cannot grow fruits (Openshaw. The third paragraph deals with the various applications for Jatropha products and specifically Pure Plant Oil and biodiesel.d). Both plant and seeds are toxic to humans and animals due to the presence of phorbol esters (Haas and Mittelbach. 111-118). 2006).1 Plant description When people use the term Jatropha. Jatropha will grow under a wide range of rainfall regimes from 200-1500mm. 2002.. Jayabalan.1 is a general description of the Jatropha tree. The species Jatropha Curcas. bush can live up to 50 years and starts to grow fruits from the 2nd or 3rd year. Its leaves have length and width of 6 to 15 cm and are shaped as is visible in Figure 2-1. Tanzania (Heller.

More detailed information on the physical and mechanical properties of Jatropha seeds is given in Table 2-1. nut and kernel provide insights on how to adapt the pressing process to Jatropha seeds. These studies also provide insight in the possibilities of using Jatropha oil for fuel purposes. This seems a reasonable when compared to claims in the comparative studies by Benge (Benge. yields will be significantly lower than for fertile lands. The fruits and seeds are shown in Figure 2-1.75-2 kg/year per plant after three to four years (Biswas. (Henning. 2006). 7 A. drying and pressing of the seeds.61 ± 26.D. 2004) and (Sirisomboom.82 67.33-2. In case of Jatropha deficiency of these nutrients leads to increased failure of seed development as was encountered during my field studies in Tanzania in February 2007.22 38. 335-345) and sunflower (35.25 ± 0.)).Sirisomboom. Physical properties Length [mm] Equatorial width perpendicular to length [mm] Breadth perpendicular to length and width [mm] Solid density [kg/m3] Bulk density [kg/m3] Mechanical properties Rupture force [N] Hardness [N/mm] Energy used for rupture [Nmm] Nut 21.D. 1996). 1-19).7 tons/acre). 2000.33 10. 7 A.52 ± 5. 119). In contrast to the overoptimistic claims of 12 tons/hectare (4. 7 A. Table 2-1 Physical and mechanical properties of Jatropha (Sirisomboom. Knowledge of the physical and mechanical properties of Jatropha is required for adequate design of machines for de-hulling. and Srikanth.84 Jatropha fruits are slightly elliptical in shape with length approximately 35 mm and contain 3 seeds on average (Sirisomboom.D.36 1040 1020 450 420 Nut Kernel 146.02 ± 1. (Heller. 2000. Variations of oil content with origin are shown in Table 2-2. The plants yield varies strongly and although it survives on marginal land.30 Kernel 15.%. 2000.72 ± 19.D. 1996.4 to 12 ton/hectare/year (Openshaw. 2006). 6 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .). The seeds weigh about 1 ton/m3.28 11. 2. 2000. Hardness values in Table 2-1 indicate that Jatropha seeds are relatively soft compared to for example rapeseed (>52.6 N/mm) (Faborode and Favier.63 ± 14. 1-8).8-6. The influence of seed quality on oil content could not be checked as the samples were provided by various parties.). The size of the seeds varies between 11-30 mm (average 21mm) in length and 7-12 mm (average 11mm) in width (2006.2 Seed properties Much research has been conducted on the composition and properties of Jatropha seeds by others ((Openshaw.03 9.44 ± 19.Chapter 2: Jatropha and its applications nutrients. 1-19).3 depending on seed orientation) (Gupta and Das.59 124. which equals 445-1335 per acre (Openshaw.). The oil content of the seeds varies with origin and growing conditions and is between 30-40 wt. For Tanzanian varieties this is approximately 38%.3-65. Kaushik. Mechanical properties such as rupture force and energy required for rupturing fruit.98 ± 6. Restrictions on how to store the seeds are linked to these properties. which makes it a high oil content seed.95 51. 7 A.42 ± 0.45 ± 0.54 7.03 69.6 tons/hectare (0. The seed yield of Jatropha has been estimated to range from 0. Claims on yields range from 0. Note that the data concerns fresh seeds and that under normal conditions the density is less than that of water.97 ± 0.8 tons/acre) calculation using the above values predict yield between 0.58 ± 0. The number of trees per hectare has been reported to vary from 1100-3300.

If planted wisely the plant can be used to prevent soil erosion or even reclaim exhausted land. The potential work that a fuel can do is determined by its energy content. In rural Africa Jatropha is mostly used as a natural fence to prevent cattle from feeding on crops for medicinal purposes and more recently for soap making by local women groups.Aderibigbe and others. Therefore PPO and biodiesel from Jatropha are discussed in more detail in paragraph 2. In addition press cake seems a valuable by-product. 2. The numbers in Table 2-3 suggest that. In Madagascar Jatropha is.1.77 MJ/litre and 37. 161-164). Most Jatropha related activities as well as this report are centred on liquid fuel. 2002. 2007.8% 36. Other options are the production of briquettes for industrial steam boilers or charcoal for household stoves. Jatropha oil shows lower energy content than conventional fuels. charcoal or biogas (Vyas and Singh.3. for the same reason.3. The bark and leaves contain dye and latex with medicinal properties for the pharmaceutical industry (Augustus.72 MJ/litre respectively. and Seiler. More detailed information on seed and oil properties is provided in Appendix A.1 and 2. reducing the difference in energy content to 8%. As mentioned earlier one drawback of Jatropha is that there are hardly any yields during the first two or three years.Chapter 2: Jatropha and its applications Table 2-2 Oil content of seeds of different origin determined by soxhlet extraction Origin Brasil Tanzania Ethiopia India Gambia Nigeria Oil content 30. liquid fuel is the highest-grade product to be obtained from Jatropha seeds. 7 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . One should think of using the press cake for the production of for example briquettes. 1997. When managed properly it has potential as a commercial crop for the production of Pure Plant Oil or biodiesel. A GTZ (Gesellshaft für Technische Zusammenarbeit) project with plant oil stoves in Tanzania indicates that the oil might also have some potential as fuel for cooking stoves. This indicates higher fuel consumption per kilometre. Jayabalan.8% 38. Nevertheless the other uses are worth mentioning as they provide insight in the total value chain of Jatropha products.1. used to support vanilla plants (Benge. 512-517). An interesting value when judging a fuel is the energy value.7% 33. As it still contains most of the nutrients it can be used to fertilize land. The press cake formed during oil production also has a wide variety of applications depending on local circumstances. judging on energy value. In Table 2-3 the calorific values of solid and liquid Jatropha fuels are compared to those of conventional fuels. When converting to the more interesting quantity MJ/litre the values for Jatropha and diesel change to 34. Possibilities for biogas production from Jatropha press cake have also been studied (2006. 223-243).2.8% 32.9% 37. 2006). One option to overcome this problem is intercropping of Jatropha with other crops like for example maize.3.7% 2. Jatropha press cake has an energy content between that of coal and wood.3 Jatropha applications Jatropha can be utilized for various purposes of which application as transport fuel is probably the most interesting one from both an economical and ecological point of view.1 Fuel In spite of all other options fuel is considered the most promising application for the Jatropha seeds.

Although much higher than the viscosity of normal diesel or biodiesel. Jayabalan.3. 159-160.091 47. 1059-1065). Table 2-4 Jatropha oil fatty acid composition (Akintayo. 1059-1065). poly aromatic hydrocarbons. the viscosity of Jatropha (34-36cST) is average compared to other plant oils such as soybean (31cST). The enormous difference in viscosity is striking. carbon monoxides. 2003 14. The main disadvantage of PPO on the other hand is its high viscosity that leads to unsuitable pumping and fuel spray characteristics. 2005.npl. 2007. Compared to conventional diesel the use of PPO in a diesel engine reduces the emission of sulphur oxides.158 44.Chapter 2: Jatropha and its applications Table 2-3 Calorific value of Jatropha oil and seeds compared to conventional fuels (Augustus.0% 31.8% 34. standards of which are between 1. Poor atomization.1** 32–42 16 37.kayelaby. and Nahar.1% 19.Mohibbe Azam.8% 7.08 43.8 ± 0.4% ± 1. 307-310.3% (18:1) (18:2) 47. smoke. Waris. The high saponification together with a low unsaponifiable mass shown in Appendix A indicates that Jatropha oil is a normal triglyceride. 2002. The acids consist of relatively long carbon chains (C14:0-C20:0) when compared to conventional diesel (C12:0-C15:0) (Knothe and Steidley. cottonseed (36cST) and sunflower (43cST) (Knothe and Steidley.0% In Table 2-5 a comparison between Jatropha and conventional diesel is made.1 PPO Vegetable oil that has not been treated apart from filtering is often referred to as Pure Plant oil (PPO) or Straight Vegetable Oil (SVO).1 ± 0. and Seiler. Jayabalan.2 44. Jatropha oil is just one of a wide variety of PPOs. The main reason to use PPO instead of converting it to biodiesel is the costs involved in the transesterification process.6% 41.6 and 6 cST.3% ± 1. 2314-2323).08 18-25. 2002. Its viscosity being 8 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .5% ± 0.co.Kandpal and Madan.8 6. 161-164. 2005.5 31.7% 6.6 Kandpal & Madan. 293-302)) Saturated acid Palmatic acid Stearic acid Unsaturated acid Oleic acid Linoleic acid (16:0) (18:0) Augustus. 1994 12. 161-164). gum formation and unburned fuel in the lubrication oil (Agarwal and Agarwal. The acid composition in Table 2-4 also hints an acidic nature of Jatropha oil. 2004.uk Calorific value (MJ/kg) 20. low volatility and insufficient fuel-air mixing can lead to combustion chamber deposits. 2005. which is harmful to rubber engine components. Parameter Jatropha seeds Jatropha press cake General purpose coal (5–10% water) * Wood (15% water) * Jatropha oil Biomass and fossil fuels Fuel oil (mexico) Crude oil Gasoline Diesel * * http://www.127 46 ** depending on residual oil content in the press cake 2. 2314-2323).1. 2007.2002 Akintayo. particle matter and noise (Agarwal and Agarwal. 1995.Augustus. The high viscosity of Jatropha oil is a result of the presence of the saturated and unsaturated acids identified in Table 2-4.8% ± 0. and Seiler.

2007.biofuel. Germany (DIN 51606) and European Organization (EN 14214) set CN at 47. FAMEs are environmentally safe.343 39. However at too high IV unsaturated molecules start to react with atmospheric oxygen and convert to peroxide which might lead to the polymerization of biodiesel into plastic-like material. 2004. and 9 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Waris. 293-302). 2005. 293-302). and Nahar. 1059-1065).732 917 ± 1 2. Waris.76 1. 49 and 51 respectively (Mohibbe Azam.3 71 ± 3 229 ± 4 103 ± 3 274 ± 3 45. Waris.52 0 11. German PPO norms are included in Appendix B to show the other specifications to comply with. Some level of unsaturated fatty acid components is necessary to prevent FAMEs from solidification.44 ± 0.1.98 ± 1. High temperatures occurring in internal combustion engines tend to accelerate this process. The CN for Jatropha biodiesel is on the high end being 52.27 35. Waris.071 80.11 10.33 12.2 Biodiesel Fuel Mineral diesel Jatropha oil 840 ± 1.36 1. The triglyceride esters of oil are changed into methanol monoesters (methyl esters). and Nahar. 2005. non-toxic and biodegradable making them a suitable transport fuel (Mohibbe Azam. This is the ability of a fuel to ignite quickly after injection.25 76. 2314-2323). The reaction equation for biodiesel production is shown in Figure 2-2. Some of the most widely used biodiesel standards from the USA (ASTM D 6751). Table 2-5 Some fuel properties of Jatropha oil and mineral diesel compared (Agarwal and Agarwal. although measures to reduce it seem necessary (Akintayo. 293-302). 2005. Biodiesel consists of fatty acid methyl esters (FAMEs) of seed oil and fats and is produced through transesterification. which indicates the degree of unsaturation (amount of double bonds). each with single fatty acid chains causing the lower viscosity of biodiesel.Chapter 2: Jatropha and its applications comparable to other plant oils makes Jatropha a potential material for fuel production. Figure 2-2 Reaction equation of transesterification (www. 2005. Parameter Density (kg/m3) Kinematic viscosity at 40°C (cST) Flash point (°C) Fire point (°C) Calorific value (MJ/kg) Carbon (wt%) Hydrogen (wt%) Nitrogen (wt%) Oxygen (wt%) Sulfer (wt%) 2. and Nahar.3.19 0.be) One of the important values for selection of FAMEs for biodiesel is the cetane number (CN).06 0 Biodiesel is a fuel type that has already been proven suitable for use in diesel engines. 307-310).31 (Mohibbe Azam. Another important criterion for biodiesel quality is the iodine value (IV). To comply with all standards the IV should be less than 115 (Mohibbe Azam. The biggest advantage of biodiesel over PPO is its lower viscosity (Knothe and Steidley. Higher values indicate better ignition quality.

293-302). 2005. Yimsuwan. 2005 points out Jatropha biodiesel to be among the twenty-six best suited FAME mixtures (from a 75 species comparison) judging CN.0 (Mohibbe Azam. and Nahar. boiling point and carbon chain length. IV and other variables like linolenic acid value.Chapter 2: Jatropha and its applications Nahar. 293-302) Jatropha biodiesel is well within range. and Pairintra. With an IV of 93. 10 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . The latter causing higher viscosity with increased chain length (Krisnangkura. 107-113). Waris. 2005. 2006. Research by Mohibbe Amam et al.

The oil is situated at different locations inside the cell. breaking and the displacement and removal of air from intermaterial voids. If Fp (induced by the screw) and Fr4 (friction between material and barrel) are larger than the sum of the other friction and pressure forces the material will be properly transported. The end pressures in a screw press can vary from 40 to 350 bar (4-35 MPA) depending on the type of press and input material. The continuous transport of material causes pressure to increase to a level needed to overcome the restriction. which need to be ruptured to free the oil.1 first explains the working principles of oil expelling in a screw press. 2005). Screw pressing is used for oil recovery up to 90-95%. 2000. 75-82).Chapter 3: The screw press 3 The screw press Chapter 3 is intended to give the reader more information about the screw pressing process.1 Working principles Mechanical pressing and solvent extraction are the most commonly used methods for commercial oil extraction. screw pressing is the most popular oil extraction method as the process is simple. The separated oil is discharged through holes provided along the press barrel. 1996. The compressed solid material or press cake is simultaneously discharged through the restriction at the end of the barrel. This pressure causes the oil to be expressed from the seeds. In spite of its slightly lower yield. 193-202) (Singh and Bargale. 1169-1176). A combined force of friction and pressure in the barrel causes the cell walls to rupture and oil to flow out of the liquid solid mixture inside the barrel. Screw presses exist with raw seed capacities ranging from 10 kg /hr up to tens of tons/hr. Therefore paragraph 3. During the pressing process oil seeds are fed in a hopper and then transported and crushed by a rotating screw in the direction of a restriction (sometimes referred to as die or nozzle). As the feeding section of a screw press is loosely filled with seed material the first step of the compression process consists of rolling. When the friction forces between screw and seed material increase beyond Fr4 slip will occur and transport will reduce or even stop. Vadke (1988) and Wang (2004). continues. Therefore most improvements made to screw presses over the last decades are based on manufacturer’s experiences and intuition rather than on the basis of physical principles (Vadke and Sosulski. 2000. globoids and nucleus represented in Figure 3-2. Finally paragraph 3. The operation principle of a screw press is rather simple to visualize although very difficult to model judging from attempts by Shirato (1971). The various screw press components have been described in detail by Ward (1976). All these elements are contained within cell walls. screw and barrel.2. 335-345). flexible and safe (Zheng and others. 2005. Singh & Bargale (2000) and Shukla (1990). 1988. Material conveyance through the press barrel is a result of the friction forces between material. while solvent extraction is capable of extracting 99% (Shahidi. As soon as the voids diminish the seeds start to resist the applied force through mutual contact (Faborode and Favier. together with other constituents like proteins. 75-82). The two main screw press designs are treated in paragraph 3. 11 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . (Singh and Bargale. 3.3 goes into more detail on the parameters affecting the screw pressing process. This force balance is represented in simplified version in Figure 3-1.

2007). and secondary consolidation consisting of material deformation and further liquid removal (Willems.Chapter 3: The screw press Figure 3-1 Forces acting on a press cake element situated between two worm shafts. 1610-1616). The extent to which consolidation occurs is mainly determined by the geometry of the extruder channel and outlet and material properties. Oil expelling is more complicated as a radial oil flow leaves the press barrel at the oil outlet. 106-110). In the 1970’s and 1980’s the flow of liquid from a solid-liquid mixture has been studied based on Terzaghi’s (1943) consolidation theory. 2007). From the point where all solids combine into a solid bed it is called consolidation (Willems. This oil flow is difficult to model and therefore theoretical models have not yet been used to improve expeller design. 1988. Filtration is expected not to be present inside a screw press (Venter and others. and Shook. is more easily understood because of a continuous material flow in and out of the press chamber. Broeck. which makes drawing up mass and energy balances relatively easy. Sosulski. Figure 3-2 Cell composition of oilseeds. Further distinction can be made between primary consolidation comprising rearrangement of the solid bed and liquid removal. 2006. Mrema and McNulty (1985) used consolidation theory to predict oil flow from rapeseed and cashew nuts in a piston ram-press. (Eggers. and Stein. thus creating a second outflow of material as is illustrated in Figure 3-3 (Vadke. 350-358) and therefore consolidation is the main process of interest here. All that comes in at the entrance comes out at a single exit. 2006. Modelling attempts Extrusion of polymers. 1990. In the late 1970’s Shirato et al extended their analysis towards continuous expression in screw 12 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Furthermore Körmendy (1974) and Shirato (1971) developed mathematical models for the mechanical separation of liquid from clay-water substances. 494-499) Expression of liquid solid mixtures generally consists of two steps: filtration and consolidation. The xdirection indicates movement towards the restriction. in this case rapeseed (Singh and Bargale. which closely resembles oil expelling. Whenever free flowing particles are present the process is called filtration.

2006. and Stein. Towards the press cake outlet the radial pressure on the mixture inside the barrel increases due to volumetric compression. Broeck. 1610-1616). 2000). who conducted early research on screw presses the pressure distribution along the shaft would look like Figure 3-4. Their model for the prediction of the amount of water separated from the claywater mixture still required prior knowledge on axial flow rate of the mixture and the pressure profile along the worm. Figure 3-4 Pressure variation based on theory of Ward for strainer press (1976) 13 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .Chapter 3: The screw press presses. The compression is induced by the worm shaft as it forces more and more material inside the same volume. The most recent attempt to model screw pressing of oilseeds was undertaken by Willems (2007) and based on the Vadke model including consolidation theory. According to Ward (1976). Vadke et al (1988) produced a more complete model to predict the throughput and the reduction in liquid content in a liquid-solid mixture in a press of unrestricted geometry. Work by others shows that the distribution might be shaped differently as presented in Figure 3-5 (Eggers. 1988. Recent studies by Willems (2007) suggest that the calculated profile measured by Eggers is more sensible than the one by Ward. Increased radial pressure increases the outflow of oil from the cells (Singh & Bargale. Results by Eggers and Ward were both intended for the ‘strainer press’ type that will be further explained in the following paragraph. This restriction makes the Shirato model unsuited for design purposes as these variables are not known in advance. and Shook. Unrestricted geometry implies that the size of the cylinder and worm can be varied freely in the model. Figure 3-3 Material flow in a section of a worm channel during oil expelling (Vadke. 494-499). Sosulski.

2000). Instead of increasing the screw diameter inserting cones on a constant diameter screw is a frequently applied alternative. Figure 3-6 Strainer press (Ferchau. For flexibility often different screw types are available or subsections can be replaced.2.Chapter 3: The screw press Figure 3-5 Pressure distribution along the screw for strainer press (Eggers. ‘Strainer presses’ are available with capacities ranging between 15 and 2000 kg/hour (Ferchau. 495-503). oil outlet and press cake restriction. The choke gap can be adjusted to influence pressure and the shape and oil content of the press cake. 3. The profile of this ‘hole cylinder press’ is shown in Figure 3-7. The gaps between the bars form an adjustable oil outlet. They mainly differ in screw geometry. The figure shows the increasing diameter of the screw in order to increase pressure in the direction of the choke. Figure 3-6 shows a schematic of the ‘strainer press’.2 Cylinder-hole press With the second press type the oil is pushed out through holes drilled in the cylinder tube. The cake is pressed out of the adjustable choke in the form of flat and often fractured flakes. and Stein. In order to 14 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . 3.2.2 The two main screw press designs Screw presses can be subdivided into two main types. 494-499) 3. Increasing pressure forces the press cake through a circular nozzle at the end of the cylinder. The screw design causes the volume displacement at the feed end to be considerably larger than at the discharge end. 1983. resembling a strainer. Broeck. namely the ‘strainer press’ and the ‘cylinder-hole press’. Spacers placed between the steel bars permit oil outlet as the pressure on the feed material increases. 1983. 2006. a hand wheel on the opposite end of the screw is typically used to adjust the choke (Khan and Hanna. For smaller machines. 2000).1 Strainer press For this first type the screw rotates in a cage lined with hardened steel bars. This adjustability allows the pressure to be optimized for different input materials. 495-503). The resulting pressure increase forces the oil through the strainer (Khan and Hanna.

the area around the nozzle (press head) is usually preheated before operation which decreases the viscosity of the paste inside the press. As no literature on the performance of ‘twin-screw extruders’ with Jatropha seeds was found they are mentioned here only for the sake of completeness and will not be treated in any more detail in this report. which is the location where maximum temperature occurs. When pressing rapeseed the temperature of the oil should not increase above 40°C. screw design and seed conditions. 1073-1080).2. due to increased resistance the temperature inside the screw press increases. The pressure level in the screw press is influenced by nozzle diameter. Pressure increases with decreasing nozzle size. should at maximum be in the range of 60 to 80°C. To prevent the oil from reaching 40°C the temperature near the die. 2000). Claims by Foidl indicate that 0. Finally the chance of blockage increases with decreasing nozzle size. One of these variations is a ‘twin-screw extruder’ which is claimed to achieve very high yields on de-hulled seeds compared to traditional single-screw presses (Dufaure and others. in order to extract as much oil as possible. Unfortunately no official results to underpin the claims by Foidl were obtained.Chapter 3: The screw press avoid blocking of the press.5% residual oil content to be possible when using CO2 as a solvent (Foidl. High phosphor content is an undesirable fuel property as it can cause deposits and clogging in engines. This effect would suggest a small nozzle. Decreasing the nozzle also has some negative affects. If on the other hand the cake outlet temperature is too low the solid content in the oil becomes unacceptable (Ferchau. 2000). 15 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Comparable adaptations are possible for the strainer based screw press. 2007). 2000). 1996). Another variation is solvent injection in an extruder. This could cause the temperature of the oil to exceed threshold values or lead to faster machine wear. because increasing temperature leads to higher phosphor content in the extruded oil. First of all the seed throughput decreases for smaller nozzle diameters.3% residual oil content can be achieved when injecting isopropanol under 300 bar pressure (Foidl. 2007). Foidl also expects 1. In case of rapeseed a rotation speed of 20 to 50 rpm is normally used because energy use is low in that case. Further advantages are flaking and boiling pre-treatments would be redundant for ‘twin-screw extruders’. Second. Thirdly the solid content in the separated oil is higher at higher pressure caused by reducing the restriction size (Ferchau. 1999. 3.3 Different designs In addition to conventional expeller design several variations have emerged over time. Figure 3-7 Hole cylinder press (Ferchau. As a rule of thumb the minimal oil content of the press cake when using screw presses should be about 5-6% (Hui.

The oil point pressure decreases while pressure build-up increases due to increased viscosity in turn drastically increasing oil recovery. This again causes the residual oil content to be relatively high (Willems. At higher speed the viscosity thus remains lower resulting in less pressure build-up.3 Parameters affecting the screw pressing process for oilseeds A number of variables need to be taken into account to optimize oil production using screw presses. By removing a part of the hard hull material less energy is needed for breaking and compressing (Zheng and others. The effect of the independent variables on oil recovery can be explained by looking at the dependent variables described below.Chapter 3: The screw press 3. For oil recovery and energy consumption pressure is more interesting to monitor. Oil recovery This is the most important dependent variable considering the goal of this report. Moisture content An optimal moisture level for oil expression is expected to exist. 2007). Temperature The friction inside the barrel generates heat which is passed on to the oil. 193-202). 3. 2000. A resulting decrease in oil content causes increased viscosity of the paste and further pressure rise. 2005. For flaxseed the optimal moisture content is expected to be around 6% (Zheng and others. As Jatropha hull contains no oil reduced amounts of hull may lead to decreased absorption of oil. For quality preservation temperature is an important parameter. thereby increasing oil recovery. Hull content Hull content is expected to affect both oil recovery and energy requirement. 130-134).1 Independent variables RPM Higher screw speed means more throughput and higher residual oil content in the press cake since less time is available for the oil to drain from the solids. For more detail on oilseed processing volume 5 of ‘Bailey’s industrial oil & fat products’ should be consulted. The friction inside the barrel generates heat which is passed on to the oil. Above certain levels the amount of phosphor and changes in acid value have a negative effect on engine performance. 193202). 16 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . 2005. A short explanation of the variables considered in this study is written down below.2 Dependent variables Variables treated in this paragraph change as a result of the independent variables treated previously.3. The optimal value for Jatropha is expected to be in the same order of magnitude Cooking Cooking causes increased cell wall rupturing thereby facilitating the outflow of oil. Restriction size When the restriction size is reduced the pressure required to overcome the restriction increases. In case of rapeseed it is a moisture level close to 7% (Bargale and Singh.3. 3.

77-84). 335-345). Oil point pressure The oil point is defined as the stage which occurs just before oil flows out of compressed solid particles (Sukumaran and Singh. 335-345). 1996. In general higher pressure will mean more oil is recovered. Low throughputs often imply high oil recovery. As this stage will always occur before oil is separated from its surrounding solids Sukumaran & Singh (1989) identified it as an important material property when studying mechanical oil expression. 17 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Should the same principles hold for an expeller it is to be expected that the oil point pressure of the material inside will increase for higher moisture level and screw speed. 335-345). The oil-point is theoretically coupled to the kernel density of an oilseed (Faborode and Favier. Size reduction and heating can be applied to increase compressibility of the seed material and thus lower the oil-point pressure (Faborode and Favier. 1996. Sukuraman & Singh found that the cell walls were still in tact at the oil-point confirming that rupturing of cell walls is not a precondition for oil expression. For the oil to be freed from the solids it has to come to the surface of the seeds and fill interparticle voids (Sukumaran and Singh. Throughput can be used to explain oil recovery and is an important variable for commercial processing.Chapter 3: The screw press Pressure When speaking of pressure radial pressure is meant in this report. The viscosity of the material to be pressed and the rotational speed will also determine the amount of energy needed. Press tests by Sukuraman & Singh (1989) using a loading piston indicate that pressure corresponding to the oil point increases with increased moisture content and with increased rate of deformation. Throughput The amount of material transported through the press per unit of time is affected by the independent variables. 1989. For rapeseed the oil point pressure was found to be 57-97 bar depending on moisture content and rate of deformation. Oil recovery is assumed to depend on the pressure exerted on the material in excess of the pressure that is required to reach the oil point. 1989. Energy requirement Energy requirement is probably related to the pressure inside the press head. Pressure is expected to be a proper measure for oil recovery. 1996. The oil can leave the cell through plasmodesmata (pores) in the cell wall (Faborode and Favier. Pressure is expected to change due to changes in all independent variables. The most significant effect can be attributed to moisture. 77-84). The oil point pressure is affected by the rate of deformation in mm/min (which can be translated to screw speed in case of an expeller) and the moisture content. The material to be pressed consists of oil bearing solids.

1. The capacity of the press for Jatropha is approximately 12kg/h.2 elaborates on tests done at Diligent Tanzania LTD. The Netherlands) to extrude the oil from the seeds in one pass. France). Standard Silica crystals were available at the faculty of Chemical Engineering. Switzerland). 18 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Hull content was varied between 60 and 100% using the custom built de-huller shown in Figure 4-2. Two different batches were supplied. The seeds underwent different pre-treatments in order to evaluate the effect on amongst others oil yields and oil quality.4. Determination of the moisture content is done by oven drying (Heraeus BR 6000.1 Experiments at Eindhoven University of Technology 4. Detailed specifications of the press geometry are tabulated in Appendix C. BT biopresser aps.1 Materials Seeds of Jatropha Curcas L were obtained from Diligent Energy Systems (Eindhoven.Chapter 4: Materials and method 4 Materials and method This chapter describes all experimental methods conducted at both Eindhoven University of Technology and Diligent Tanzania LTD. one harvested in July 2005 and the other one in July 2004. Right: front view inside the press head. The prepared samples each weighing about 2 kg were either sealed for storage or instantly fed into the screw press. The variable speed drive is also used for readings on power consumption of the electric motor. was used.1. 4.% (dry basis).1 the experiments conducted at the university are described. Pressures during testing reached up to 300 bars at temperatures of over 140 ˚C. Germany) just before the seeds are fed to the screw press in order to minimize errors. This small capacity press was designed for smaller seeds like rapeseed and sunflower but proved to run properly on the bigger Jatropha seeds. The oil content of the seeds. was found to be 37. Figure 4-1 Left: BT50 test setup at Eindhoven University of Technology.8 wt. Rueil-Malmaison.1 kW electric motor. Denmark) was made available by Diligent Energy Systems and FACT Foundation (Eindhoven. In paragraph 4. Figure 4-1 shows the BT 50 test setup. 4.1. Paragraph 4. The Netherlands). which was not optimized for Jatropha. determined by soxhlet extraction with petroleum ether. Hanau. Middle: situation of thermocouples and pressure sensor (blue wire). Buchs.2 Experimental setup A Danish screw press (BT Bio Presse Type 50. The origin of the seeds is the northern region of Tanzania around Arusha. Schneider Electric SA. Dybvad. Screw RPM is directly regulated between 0-70 RPM by a variable speed drive(Altivar 31. Petroleum ether (boiling range 40-60˚C) was bought from (Fluka Riedel de Haën. Separation of hull and kernel had to occur manually. The restriction size at the press cake outlet can be varied by placing different sized nozzles. Further details on the soxhlet extraction are given in paragraph 4. During press tests only a single screw geometry. The screw is powered by a 1.

1. 19 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Kistler Instruments AG. To prevent jamming the press head is heated to 70-80˚C using a hot air pistol. Power consumption was recorded using one of the Altivar 31 voltage output signals. In addition the time required for a batch to pass through the press was measured. It was difficult to determine the instant at which the pressing finished. All moisture contents were determined according to DGF standard method B-I 4 (DGFEinheitsmethoden. National Instruments. therefore the time measurement was started when the pressure showed a steep increase after seeds were fed and stopped once pressure drastically declined while the press was almost empty. Temperature is measured at 4 positions along the press cage (type J Thermal Insulated thermocouple. 193-202). 1988. The output signal ranged from 0-10V where 10V indicates twice the nominal power of the 1. Switzerland). Jatropha seeds with increased different moisture content were obtained by equilibrating the seeds for one week in a sealed environment with a fixed amount of water inside plastic bag. A similar method was used in other studies on mechanical expression of oil seeds (Vadke and Sosulski.1kW electric motor (Altivar 31 user manual). 4. The silica was oven dried overnight at 103˚C. 2005. The sensor signal is amplified into a signal in the range of 0-10V by a charge amplifier (5073 Industrial Charge Amplifier Manufacturing. 2002). The first is located at the press cake exit. Switzerland) at the point of maximum expected pressure. All data was digitally recorded using a PCMCIA card (NI4350. To decrease the moisture content below the natural value of around 7% a similar method was used with silica crystals. While continuing heating small amounts of seeds are fed into the hopper. Crude oil is collected at the oil outlet and press cake exit on the left (see Figure 4. The radial pressure is measured (6152AAA Mold Cavity Pressure Sensor.Zheng and others. 1169-1176.3 Experimental procedure The description of a typical experiment is used to explain the experimental procedure. Kistler Instruments AG. the second at the point where maximum temperature is expected and the third & fourth at different positions along the oil outlet as indicated in Figure 4-3.1 middle) in the shape of small pellets. A graph in appendix D shows maximum power measurement while the press is jammed. A sample size of 2kg was chosen as this allows stable measurements and simplifies the calculation of throughput.Chapter 4: Materials and method Figure 4-2 Hand powered de-huller designed together with Gerard van Hout from the workshop at Mechanical engineering. which coincides with the point of maximum temperature. United States) and ‘Virtual Bench Logger’. Themocoax. Netherlands).

Schleicher & Schuell Microscience GmbH. After leaving the oil to settle for a week relative amount of foots in the crude oil could be determined. Hanau. The petroleum ether was heated to 100°C to ensure sufficient solvent flow rate. Haan. Standard cellulose extraction thimbles (603 Cellulose Extraktionshülsen. After extraction the obtained oil sample was oven dried at (103 ± 2) °C for 5 hours to remove remaining solvent. Figure 4-4 Soxhlet extraction setup at Eindhoven University of Technology 20 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .Chapter 4: Materials and method Point 1 Point 4: press cake exit Point 2: oil outlet Point 3: maximum temperature and pressure Figure 4-3 Location of measurement devices After pressing a sample of the press cake was analyzed on oil content with Soxhlet extraction according to DGF standards (DGF-Einheitsmethoden. Germany).4 Analysis Oil content Prior to determining the oil content of seed or press cake the sample was dried at 103˚C for 5 hours in a drying oven (Heraeus BR 6000. was determined by comparing the weight of the extracted oil to the total weight of the sample prior to extraction. Whatman group. Dassel. The oil contents are compared on a dry basis for ease of comparison.1. 2002). Retsch GmbH & Co. Germany). expressed in g/100g. Germany) were used to prevent solid particles from entering the oil sample. Oil content was determined through overnight Soxhlet extraction with petroleum ether with the setup in Figure 4-4. After drying the sample was ground together with a small amount of petroleum ether for 20 seconds at 25Hz in a ball mill (MM301. 4. The oil content.

The capacity of the Sayari for sunflower is around 140kg/hr (Peter Chisawillo. Oil analysis Analyses of some oil samples on phosphorous content and acid value was out contracted to consulting engineers (Ingenia.2. 2007. 21 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .%.Peter Chisawillo. Director Intermech) and the minimum processing capacity is 70kg/hr (Operator’s instruction manual). A simple transmission made of sprockets and chains reduces the rotational speed of the press worm to 55 RPM. The oil content of the seeds.Chapter 4: Materials and method Moisture content A drying method according to DGF standards was used to determine the moisture content of the seeds before pressing.1 Materials Seeds from Jatropha Curcas L were obtained from Diligent Tanzania (Arusha. Production is subcontracted to three workshops located close to the Vyahumu office. Tanzania). determined by soxhlet extraction with petroleum ether and hexane.2. Pressures inside the press can reach values up to 150 bars at temperatures of around 110˚C.1. including pictures showing the assembly of the Vyahumu press is shown in Appendix F). The samples consisted of crude Jatropha oil that had been subjected to different cleaning methods. Analysis of the samples was done by ASG Analytik-Service Gesellschaft mbH (Neusäss. depicted in Figure 4-5. Morogoro. Samples of approximately 5 grams were weighed and put in the oven (Heraeus BR 6000. 2007) the capacity for Jatropha should be about 70-80 kg/hr. The four cones mounted on the screw (see Figure 4-5 right side) enable pressure to build up by gradually reducing the clearance between screw and press cage. Marelli Motori.2 Experimental setup Oil expelling tests were performed using a Sayari expeller (Vyahumu Trust. The Sayari press. due to superior equipment. Hanau. were found to be 38 wt. After drying the weight was again determined. Germany) for 5 hours. 4. The press is powered by a 5. Eindhoven. The drying method is similar to the one used before soxhlet extraction. Tanzania).5 Accuracy and reproducibility of experimental results Results obtained at Eindhoven University of Technology are more accurate than the ones from Tanzania. Repetition of measurements would have increased reliability and accuracy.2 Experiments at Diligent Tanzania LTD As official descriptions of some tools were not available pictures and additional information have been included in Appendix E. one of which is Intermech (More information on the workshop and Sayari expeller. is now a locally produced screw press that was initially designed to process sunflower seeds although suitability to press Jatropha is mentioned in the ‘Operator’s instruction manual’.1. 4. known under the name Sundhara.Shila. originates from India it was adapted to Tanzanian standards by engineers from GTZ in the early 90’s.% and 39 wt. High pressure and friction inside the press cause significant heating of the press during operation. Germany) according to DIN methods. All the seeds used for this research originated from the same village Engaruka located 180km North West of Arusha in order to create homogeneous samples.% successively (dry basis). Italy) rotating at 1500 RPM.7 wt.5 kW asynchronous generator (MA 132 SA 4. The Netherlands). 4. More information on accuracy is written in paragraph 5. According to Vyahumu and Intermech (Lehada C. Soxhlet extraction can be repeated at 1% accuracy according to DGF standards. The moisture content of stored seeds during the testing period was determined to be 6. 4. Although the original design of this strainer press. Press experiments were only conducted once due to time constraints. Tests at Diligent Tanzania LTD showed similar results.

A charge amplifier (5073 Industrial Charge Amplifier Manufacturing. The normal seed batch size for measurement was 10-20kg because it takes some time for the process to stabilize. All data was digitally recorded using a PCMCIA card (NI4350. Zheng. Switzerland). It was difficult to determine the instant at which the pressing finished. After leaving the oil to settle for a week relative amount of foots in the crude oil could be determined. After pressing a sample of the press cake was analyzed on oil content with Soxhlet extraction according to DGF standards (DGF-Einheitsmethoden. Wiesenborn et al 2004) reported to have used similar methods. 1988. with (1) feeder.7% by sun drying. United States) and ‘Virtual Bench Logger’.3 Experimental procedure The description of a typical experiment is used to explain the experimental procedure. In addition the time required for a batch to pass through the press was measured. Kistler Instruments AG. A thermocouple (type J Thermal Insulated thermocouple. National Instruments. Crude oil was collected at the oil outlet (2) and press cake exits the machine at the adjustable cone (3). The fourth cone. located at the press cake outlet. Kistler Instruments AG. Before starting the measurement the press required preheating to 60-70˚C during approximately 20 minutes of normal operation using seeds. Left: Sayari press at Diligent Tanzania premises. 4. Jatropha seeds with increased moisture content were obtained by equilibrating the seeds for one week in a sealed environment with a fixed amount of water inside a plastic container. After preheating seeds or press cake were inserted in the feeder (1) on the right side of the press shown in Figure 4-5. press cake or a mixture of the two. 94 and 98mm (4) from right to left. Right: Worm inside Sayari screw press showing 3 cones with diameters 90. The remaining press cake was fed back into the press and the whole process repeated for a second time.Chapter 4: Materials and method Figure 4-5. Themocoax. As this method depends 22 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . therefore the time measurement was started when the pressure showed a steep increase after seeds were fed and stopped once pressure drastically declined while the press was almost empty in order to generate comparable results. Other studies on mechanical expression of oil seeds (Vadke and Sosuslski. is adjustable by a hand operated turning wheel on the far left. Moisture content was reduced below the natural 6. 2002). The radial pressure was measured at the same cone (6152AAA Mold Cavity Pressure Sensor.2. Switzerland) transformed the sensor signal to a signal in the range of 0-10V. Netherlands) located at the third cone (see Figure 4-5 (4)) measured temperature changes. (2) oil outlet. (3) press cake outlet.

Kern & Sohn GmbH. These deviations from the standard method lessen the accuracy of the results.2. Samples of ± 10 grams were used instead of the prescribed ± 5 grams in order to reduce the effect of measurement errors. The oil content. expressed in g/100g. A correction was applied afterwards based on drying of representative samples. 1001125) instead of extraction thimbles. Samples of approximately 5 grams were weighed and put in the oven (Nevica. see Appendix E for picture and drying profile) for 3 hours. The method differed from DGF standards as some of the required equipment was not available in Arusha. Due to the simple oven design a drying profile had to be determined in order to keep the temperature in the acceptable range of 80-100˚C. Oil and foots levels were marked and refilling with known quantities of water resulted in an approximation of the actual volumes. Gottl. However comparative tests for Jatropha with petroleum ether showed comparable results with hexane the values were 38% and 39% respectively. Figure 4-6 Oil samples left for settlement 23 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Moisture content A drying method according to DGF standards was used to determine the moisture content of the seeds before pressing.4 Analysis Oil content Oil content was again determined through soxhlet extraction with hexane.001g. 4. 0.01g instead of 0. A 1000ml graduated measuring cylinder with 5ml accuracy was used for the water refilling. Balingen-Frommern. The samples were not dried before extraction. Hexane was used as solvent instead of petroleum ether. was determined by comparing the weight of the extracted oil to the total weight of the sample prior to extraction. The setup differed from DGF standard on the following aspects: • • • • • • A mortar and pestle were used to grind samples prior to extraction instead of the ball mill. The electrical balance (440-33N. designed for domestic use. Samples were wrapped in normal filter paper (Whatman qualitative filter paper 125mm.Chapter 4: Materials and method heavily on weather conditions the drying time was not constant. Foots The relative amount of foots in the crude oil were determined after one week of settling in plastic bottles (see Figure 4-6). Cat No. Germany) did not meet accuracy requirements. Due to the less accurate soxhlet setup in Tanzania samples were transported to the Netherlands for accurate extraction. Resulting moisture contents were determined by oven drying.

The accuracy for measuring oil and foots levels was estimated at ±50ml.5 Accuracy and reproducibility of experimental results Repetition of each pressing experiment would allow proper judgment of accuracy. Due to limitations in both means and time repetition was seldom possible for the press tests executed in Tanzania.Chapter 4: Materials and method 4. To what extent this affects pressure measurements could not be determined. 24 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . For example oil content found to be 37% means that the actual value lies between 35-39%. In addition to the less accurate mounting of the pressure sensor its position relative to the third cone changes up to 1cm when changing restriction size. due to inaccurate reading of the marked bottles. Results on oil yield. Less accurate mounting of the pressure sensor in the Sayari expeller compared to the BT50 caused pressure values to deviate ±20% instead of 10% as will become clear from the results section 5. temperature are therefore meant to identify trends and are assumed to deviate up to ±2 and ±5%.2.1. This comes down to deviation in the range of 5-15% depending on sample size.

2007). For the BT 50 all tests with respect to seed pre-treatment were conducted at standard conditions which are 49 RPM. 100% hull and a restriction diameter 9mm. The standard settings for the Sayari expeller were 55 RPM. The following paragraphs provide an overview of the most relevant test results and an explanation of what causes the observed differences. These settings were selected as standard setting because they ensured smooth running of the presses. All tests in the BT50 press were single pass. Finally the experimental results are used to optimize the pressing process. 2006. Table 5-1 Overview of the tested variables for the two different presses BT50 Press settings RPM Restriction Seed pretreatments Moisture content Dehulling Torrefaction Heating in water Crushing X X Sayari X X X X X X X X X X The effect of a pre-treatment on the pressing process is best explained by the behaviour of pressure and temperature inside the press (Eggers.8 mm 9 mm Figure 5-1 Schematic representation of the restriction shape of the Sayari expeller (left) and the BT50 (right). meaning that only seeds were pressed and the press cake was not fed back for a second run. 494499. Most tests in the Sayari expeller were dual pass tests.8mm. An overview and explanation of the exact values for speed.Vadke and Sosulski. Broeck. 1. Single pass and dual pass are indicated in the figures with Sayari results.Willems. The restriction sizes of the two presses schematically indicated in Figure 5-1 refer to different geometries and can therefore not directly be compared. 25 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . moisture content etc is given in Appendix G. 1988. The test settings for the two different press types that were used are as follows. 100% hull and a restriction gap size of 1. Because of the higher residual oil content of the press cake from the Sayari expeller a second press run was worthwhile. The white circle on the left side represents the adjustable cone of the Sayari expeller. Increased oil yield can be achieved either by increasing the amount of crude oil or by reducing the relative amount of foots in the oil.Chapter 5: Results and discussion 5 Results and discussion The objective of this chapter is to show the effect of various seed pre-treatments and press settings on the oil yield of mechanical expression of Jatropha seeds. and Stein. 1169-1176. The variables included in this research are listed below in Table 5-1.

307-310. 26 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Only pressure values after a sudden rise. which means that a value of for example 37% is between 35-39%. The available recording equipment lead to the selection of thermocouples that were suboptimal for the temperature range and therefore the variation in temperature is assumed to be 5%. Thermocouple calibration in a thermostatic bath showed an accuracy of 2. indicating pressure build up inside the press. Starting time was identified by a sudden increase in pressure.Ruud van Eck.6kg of oil are recovered. Timing errors are assumed to be within 30 seconds leading to 5% inaccuracy for large batches (BT50 tests with 2kg batches and Sayari tests with 20kg batches) and 10% inaccuracy for small batches (BT50 tests with 1kg batches and Sayari tests with 10kg batches). In spite of the Kistler 6152AAA pressure sensor being 1% accurate a 10% error is used to take the fluctuations into account. 2004. The main points concerned here are data selection and error margin determination Oil recovery The oil recovery graphs were obtained by comparing the residual oil content in the press cake to the initial oil content in the seeds using the following formula (see Appendix H for derivation):   Ws     1 − Ws    Y = 1−    Wo     1 − Wo     with Y= oil yield or oil recovery (%oil/oil) Wo= Oil content in original seed (g/100g) Ws= Oil content in press cake sample (g/100g) The oil recovery is expressed as the recovered fraction of the initial amount of oil. were included. Throughput Throughput was determined by dividing the batch weight by the time required for pressing the batch.Chapter 5: Results and discussion 5. Graphs showing the initial measurement output are shown in Appendix I to support this choice. The accuracy of the soxhlet extraction is stated as 1% (DGFEinheitsmethoden. 2006) this means 32. Taking into account the density of 0. Time recording is slightly inaccurate as it is hard to determine at what moment the last material exits the press. Temperature The temperatures as indicated in figures similar to Figure 5-2 b represent the average value at the same axial location as the pressures. Pressure The graphs indicating pressure dependence were compiled using average pressure values. Due to the rotating screw the pressure occurring in the press head is very turbulent.2 litre. So knowing that the Jatropha seed contains 37wt. 239248.1 Data processing This paragraph briefly explains the coming about of the graphical representations in the following paragraph.92kg/l (Akintayo. The error bars indicate a ± 2% fixed error.% oil and assuming an imaginary recovery rate is 80% this means that from 100kg seeds 29. 2003. The other 1% is attributed to variations in the oil content in the press cake when it leaves the machine.5% in the temperature range of 90-100°C and less than 1% for lower temperature (see Appendix D).Pramanik. 2002).

2 Influence of RPM BT50 The graph in Figure 5-2 a shows a reduction in oil recovery for increased rotational speed.7l/hr at 70RPM. As the accuracy was not given by the producer a 5% error was assumed for both graphs. The negative slope of the upper line indicating energy input per litre in Figure 5-2 c shows 27 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) . This effect is largely explained by the increasing throughput indicated in the same figure. Under uni-axial compression the oil-point pressure increases with increasing rate of deformation (Sukumaran and Singh. Oil recovery 100 Throughput 14 140 Pressure Temperature 140 12 80 120 120 Oil recovery (%oil/oil) Pressure (bar) 60 Throughput (kg/hr) 10 100 100 8 80 80 40 6 60 60 4 20 2 40 40 20 20 0 20 30 40 50 60 70 0 0 20 30 40 50 60 70 0 Screw speed (RPM) Screw speed (RPM) Figure 5-2 a Influence of screw speed on oil recovery and throughput for the BT50 Figure 5-2 b Influence of screw speed on radial pressure and temperature for the BT50 The energy requirement to press the seeds is represented in Figure 5-2 c. The lower pressure is a second cause for the oil recovery to be lower for higher speeds. An approximation of the energy requirement for the BT50 made prior to testing is shown in Appendix J. 1989. Only the pressure exerted in excess of the oil-point pressure is assumed to result in oil to be recovered from the compressed seeds. Increased throughput means reduced residence time and thus less chance for the oil to flow from between the solid material. 5. Higher rotational speed requires more power and at the same time reduces processing time. Applying the same assumption to screw pressing would predict lower oil yield at higher screw speed.Chapter 5: Results and discussion Energy input The graphs for energy input were computed from averaged power output readings of the Altivar 31 variable speed drive converted into kJ/kg and kJ/l values. The rotational speed of the worm inside the screw press could be considered analogue to the rate of deformation under uni-axial compression. 77-84). The increase in temperature for higher speed could be explained by increased friction heating or simply be due to the fact that the tests were executed successively from low to high speed and the press had not yet reached equilibrium temperature. Unfortunately the energy input could not be measured during the experiments carried out on the Sayari expeller in Tanzania. Combining throughput and oil recovery reveals that the oil production capacity varies from 4. The higher residual oil content in the material ensures that the viscosity of the paste remains relatively low and therefore pressure build-up is also lower at higher speed as can be seen in Figure 5-2 b. Temperature differences can be considered insignificant for changes in rotational speed. A third cause for decrease in oil recovery with increasing screw speed can also be an increase in oil-point pressure as studies by Sukurman & Singh (1981).2l/hr at 28RPM to 8.

Energy input (kJ/l) Oil recovery 100 Throughput 16 14 Pressure 160 150 140 130 12 Temperature 140 80 120 120 Oil recovery (%oil/oil) Throughput (kg/hr) 110 60 10 8 Pressure (bar) 100 100 90 80 70 60 50 40 30 20 10 0 6 8 10 12 40 80 40 6 4 60 20 2 0 6 8 10 12 0 20 0 Restriction size (mm) Restriction diameter (mm) Figure 5-3 a Influence of restriction size on oil recovery and throughput for BT50 press Figure 5-3 b Influence of restriction size on temperature and pressure for BT50 press Because the throughput for 7mm was not measured it is also not possible to calculate the required energy input. The two data points in Figure 5-3 a incline to a slight decrease of throughput with reduced restriction size due to the balancing effect of increased 28 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) . Energy input per kg seed 400 Eenrgy input per liter crude oil 400 350 350 300 300 Energy input (kJ/kg) 250 250 200 200 150 150 100 100 20 30 40 50 60 70 Screw speed (RPM) Figure 5-2 c Influence of screw speed on energy requirement for the BT50 expressed per kg and per litre crude oil 5. Judging from the figures the higher oil recovery for smaller restriction can directly be attributed to increased pressure and friction.3 Influence of restriction size BT50 Figure 5-3 a and Figure 5-3 b show the influence of the choke opening on pressure and residual oil content for the BT50 press. Decreasing the choke opening increases the resistance to flow at the end of the extruder. It is difficult to draw conclusions on the effect on throughput as only two data points are available.Chapter 5: Results and discussion that the energy requirement is most strongly affected by the reduction in processing time. The pressure required to overcome this resistance significantly increases judging from Figure 5-3 b.

Although limited data is available on the effect of restriction diameter on energy requirement for the BT50 Figure 5-3 c suggests a reversed dependence between restriction size and energy input. Input energy (kJ/l) 250 250 Oil recovery single pass 100 Throughput single pass 110 120 Pressure single pass Temperature single pass 80 70 100 80 90 100 60 50 40 60 80 80 40 70 60 30 40 20 20 10 0 0 1 2 3 4 5 60 20 50 0 0 1 2 3 4 5 40 0 Restriction (mm) Restriction (mm) Figure 5-4 a Influence of restriction size on oil recovery and throughput for Sayari press Figure 5-4 b Influence of restriction size on temperature and pressure for Sayari press The supply of input material is assumed to stay more or less constant resulting in increased compression close to the press cake outlet as is indicated by the pressure development in Figure 5-4 b. Energy input per kg seed 400 350 300 Energy input per liter crude oil 400 350 300 Input energy (kJ/kg) 200 150 100 50 0 8 10 12 200 150 100 50 0 Restriction diameter (mm) Figure 5-3 c Influence of restriction size on energy requirement for the BT50 expressed per kg and per litre crude oil Sayari As can be seen in Figure 5-4 a reducing restriction size clearly results in an increase in the amount of oil recovered from the seeds. For a second pass the shape of the oil recovery graph is expected to be similar in shape although less steep. The expected reduction in steepness is due to being closer to the theoretic maximum oil recovery at smaller restriction sizes after the first pass. The effect of restriction size was only observed for a single pass as this already shows how this variable affects oil yield. Obviously when decreasing the restriction size the mixture inside the press experiences more resistance when exiting the press cake outlet. The increase in throughput with decreasing restriction size 29 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) Oil recovery (%oil/oil) Throughput (kg/hr) Pressure (bar) .Chapter 5: Results and discussion resistance and better filling of the press chamber.

This pressure drop seems an appropriate explanation for the rapid decline in oil recovery in Figure 5-5 a.Chapter 5: Results and discussion can appear conflicting at first sight. Although oil recovery is highest at smaller restriction size there is an operational limit.) Figure 5-5 a Influence of moisture content on oil recovery and throughput for the BT50 press Figure 5-5 b Influence of moisture content on temperature and pressure for the BT50 press 30 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) Oil recovery (%oil/oil) . The effect of moisture content on oil recovery is on the other hand considerable. Only part of the decrease in oil recovery with increased moisture content can be attributed to the increase in throughput. The suggestion by Willems (Willems. Although oil recovery is optimal for a moisture content of 2. Higher moisture content might cause increased deformability and reduced rupture fore as suggested by Singh & Bargale (2000). In addition emulsification or plasticizing effects occurring in the pressed material at higher moisture levels reduce the viscosity and thereby decrease pressure build-up. The tightest setting of 0. although the decrease is less than with increasing screw speed. The cooling effect of the cold material entering the press might in this case counterbalance heat generation or even lead to faster cooling than is the case after normal experiments. When comparing Figure 5-5 a to Figure 5-2 a it is apparent that with increasing moisture content the increase in throughput is smaller and the decrease in oil recovery is bigger. The increase in temperature with decreasing restriction size is easily explained by increased resistance and pressure. temperature and pressure are not indicated in Figure 5-4 a and Figure 5-4 b. The strong decrease in pressure visible in Figure 5-5 a hints another cause for the reduction in oil recovery. 75-82.Zheng and others. 5.4 Influence of moisture content BT50 The results of press tests for varying moisture content using the BT50 press are shown in Figure 5-5 a to Figure 5-5 c. A more plausible explanation is a cooling effect of the input material.) Moisture content (wt% w. 2000. Figure 5-5 a shows that with increasing moisture content the throughput decreases. For this reason corresponding values for throughput. The amount of water present in the seeds however seems insufficient to induce the temperature drop by more then 50%. Oil recovery 100 14 Throughput 350 Pressure Temperature 120 300 80 12 100 250 Throughput (kg/hr) 10 60 8 Pressure (bar) 80 200 60 40 6 150 4 20 2 100 40 50 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 15 0 2 4 6 8 10 12 14 0 Moisture content (wt% w.14% this might not be the optimal processing conditions as the oil temperature reaches high values that could lead to changes in the composition of the oil as is the case for rapeseed (Singh and Bargale. 2005. Lower friction and pressure build-up due to the reduced viscosity reduce heat generation at higher moisture content.b. 193-202).6mm exceeded the operational limit of the Sayari press resulting in severe jamming.b. which directly points out that changes in throughput caused by varying moisture content are an insufficient explanation for the occurring changes in oil recovery. 2007) that during expression water can only be removed by evaporation brings up another contribution to the cooling effect. It might be contributed to better filling of the voids inside the worm channel at higher pressure therefore enhancing the press’s performance.

Judging from Figure 5-5 b and Figure 5-6 b for moisture levels between 6 and 9% the temperatures in the Sayari press are about 15% lower than for the BT50. In the same moisture range the pressure measured in the BT50 is approximately two times the value measured in the Sayari expeller. as was the case for the BT50. The slopes of both temperature lines are comparable. This example directly shows that energy requirement is not expected to be a limiting factor. Temperatures and pressures are also less aligned compared to the previous figures.Chapter 5: Results and discussion Judging from Figure 5-5 c the energy requirement per kilogram of input material reduces for increasing moisture content. material and design of cage and screw. Due to the design of the Sayari expeller it is much more difficult to execute steady test runs. From a cost perspective it is not desirable to decrease the energy use by increasing the moisture level in the seeds as the costs of three times as much input material per litre are much higher than the savings of using three times less energy. friction. An explanation could lie in the multiple stage compression in the Sayari expeller. More seeds per litre mean longer processing time per litre. but the absolute temperature value is largely determined by the combined effect of pressure. This results in the decreased slope of the upper line in Figure 5-5 c at higher moisture levels. Test results from the Sayari expeller in Figure 5-6 b show a drastic decrease in pressure with moisture content. The effect of moisture is less than for the BT50. Oil recovery after dual passing shows a slight decrease for increasing moisture content. Although data on throughput seem inaccurate the increase in throughput with moisture level is plausible when compared tot the BT50 results in Figure 5-5 a. In addition conditioning of the seed moisture level was less accurate due to less sophisticated instruments. However at higher moisture content more seeds are required judging from Figure 5-5 a. The uneven distribution of data points in Figure 5-6 a is the result of this unsteady operation and imprecise conditioning. This shows that changes in pressure are an indicator for changes in temperature. and to a lesser extend temperature. Energy input per kg seed 900 800 700 Energy input per liter crude oil 900 800 700 Input energy (kJ/kg) 500 400 300 200 100 0 2 4 6 8 10 12 14 500 400 300 200 100 0 Moisture content (wt%) Figure 5-5 c Influence of moisture content on energy requirement for the BT50 expressed per kg and per litre crude oil Sayari The effect of moisture content was tested for dual passing in the Sayari expeller. Input energy (kJ/l) 600 600 31 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .

5 0 40 Temperature dual pass 100 Pressure (bar) 60 Moisture content (wt% w.0 9.5 40 Moisture content (wt% w.0 8. 1996.b.5 Influence of hull fraction BT50 As the oil-point pressure is lower for softer oilseeds the same will probably hold for dehulled seeds compared to normal seeds as the bulk material becomes softer (Faborode and Favier.) Moisture content (wt% w.) Figure 5-6 c Influence of moisture content on temperature and pressure for Sayari press (dual pass) 5.5 7.5 9.5 8.0 7. Pressure dual pass 80 70 80 60 50 40 30 20 20 10 0 6.0 8.0 7. What happens when de-hulling the seeds is that the material inside the press becomes fattier and due to a difference in shape friction between the mixture and the barrel drops below the friction between the mixture and the worm. Based on this assumption a higher oil recovery is expected although Figure 5-7 a clearly shows the opposite. When the hull content is reduced to 66% of its original value causes slip in the feed section and the solid-oil mixture starts to stick to the worm.b.0 20 40 40 80 70 20 60 50 0 6. The result is a sharp decrease in throughput (Figure 5-7 a) which can even drop to zero.5 9. Due to this unintended preheating of the press cake comparison between pressing seeds and press cake is difficult.5 7.b. 335-345).5 9. The higher temperature is the result of repressing the press cake while it is still hot (about 60-80°C).5 8. The Temperature (degreed Celsius) 32 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) .0 7.0 8.0 6.5 0 10.0 9.) Figure 5-6 a Influence of moisture content on oil recovery and throughput for the Sayari press Figure 5-6 b Influence of moisture content on temperature and pressure for the Sayari press (single pass) When comparing Figure 5-6 b and Figure 5-6 c one can see the difference between pressing seeds (single pass) and pressing press cake (dual pass).5 7.Chapter 5: Results and discussion Oil recovery dual pass Oil recovery single pass 100 Throughput of seeds average single and dual pass 150 140 Pressure single pass 90 80 70 60 Temperature single pass 80 80 130 120 Oil recovery (%oil/oil) 60 Pressure (bar) 110 100 90 Throughput (kg/hr) 60 50 40 30 20 10 0 6. What stands out is that the temperature at dual pass is higher while the pressure is lower.5 8.0 9.

Below a certain threshold value between 66-80% hull the press performance is sharply reduced due to deteriorating mixture conveyance. Nevertheless the results in Figure 5-8 a and Figure 5-8 b indicate a dependence on hull fraction Energy input (kJ/l) 700 33 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) 80 12 130 Throughput (kg/hr) . Oil recovery 100 Throughput 14 Pressure 150 140 temperature 100 90 80 70 60 50 40 30 20 10 0 120 Oil recovery (%oil/oil) 10 60 8 110 100 Pressure (bar) 90 80 70 60 50 40 40 6 4 20 2 30 20 0 65 70 75 80 85 90 95 100 0 105 10 0 60 70 80 90 100 Hull content (wt%) Hull content (wt%) Figure 5-7 a Influence of hull content on oil recovery and throughput for BT50 press Figure 5-7 b Influence of hull content on temperature and pressure for the BT50 press The required energy input decreases for lower hull content as indicated by Figure 5-7 c. Throughput reduced to zero during the experiment thereby causing the deviant result. The combined effect of a decrease in power consumption (expressed in Watt) and an increase in throughput at comparable oil yield is increased energy efficiency of the process (less kJ/litre) when hull fraction is reduced to 80%. Khan & Hanna (1983) suggest the use of a low friction shaft combined with higher friction cage bars for softer seeds like de-hulled Jatropha. The data points for 66% hull content are too high due to slip and rotation movement of the solid-oil mixture instead of horizontal conveyance. Difficulties experienced during testing demonstrate that in standard screw presses a reduction of the hull fraction is applicable up to 80% of the original hull fraction.Chapter 5: Results and discussion same problem was also brought forward by Ward (1976) when dealing with high oil content seeds. Decrease in pressure and temperature are as before due to reduced friction (Figure 5-7 b). Energy input per kg seed 1100 1000 900 800 Energy input per liter liter crude oil 1000 800 Energy input (kJ/kg) 600 500 400 300 200 100 0 65 70 75 80 85 90 95 100 600 400 200 0 105 Hull content (wt%) Figure 5-7 c Influence of hull content on energy requirement for the BT50 expressed per kg and per litre crude oil Sayari The test results on hull content for the Sayari expeller were only limited.

Only dual pass values are given as they tend to show less variance than single pass values as was also observed in Figure 5-6 a.7% moisture) during 30 minutes at various temperatures. The effect of preheating of the seeds before pressing was tested by preheating standard seeds (6. 1996. Oilrecovery 100 Throughput 14 Pressure 140 Temperatur 160 140 120 120 Throughput (kg/hr) 10 60 8 100 Pressure (bar) 100 80 80 60 40 20 0 100 120 140 160 180 40 6 60 4 20 2 40 20 0 100 120 140 160 180 0 0 Seed temperature (degrees Celsius) Seed temperature (degrees Celsius) Figure 5-9 a Influence of seed input temperature on oil recovery and throughput for the BT50 press Figure 5-9 b Influence of seed input temperature on temperature and pressure for the BT50 press 34 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries Temperature (degrees Celsius) 80 12 Oil recovery (%oil/oil) Temperature (degrees Celsius) . The effect on oil recovery and throughput is negligible. The inner structure of the Jatropha seeds was not visibly affected after heating.6 Influence of seed preheating BT50 Heating of the input material can be applied to increase compressibility of the seed material and thus lower the oil-point pressure (Faborode and Favier. Oil recovery dual pass 100 Throughput average single and dual pass Pressure dual pass 90 Temperature dual pass 100 90 80 80 70 80 80 60 Oil recovery (%oil/oil) Throughput (kg/hr) 70 70 40 Pressure (bar) 60 50 40 30 20 60 50 40 30 20 60 20 10 0 60 70 80 90 100 50 10 0 60 70 80 90 100 0 Hull content (wt%) Hull content (wt%) Figure 5-8 a Influence of hull content on oil recovery and throughput for Sayari press Figure 5-8 b Influence of hull content on temperature and pressure for the Sayari press 5. Moisture content of the seeds was not determined before pressing. Increase in temperature can be explained through a combination of the heated input material and minor increase in friction as the material is dryer for higher preheating temperatures. 335-345).Chapter 5: Results and discussion comparable to the BT50 press. The influences on oil content and temperatures inside the press are depicted in Figure 5-9 a and Figure 5-9 b.

3% Oil recovery ress head temperature [° [° [bar] Throughput [kg/hr] Energy [kJ/l] P Oil temperature C] Pressure C] 79.0% 11. The reduction in energy use would not balance the energy needed for preheating and is thus not advisable for Jatropha seeds. 335345).1% 6.8 The effects on foots A striking characteristic of the crude Jatropha oil obtained by screw pressing is the very high amount of foots or sediments in the oil. * Samples were dried to certain moisture content before pressing.4% 5.Chapter 5: Results and discussion The assumption of minor friction increase is supported by Figure 5-9 c.7 9.7%. The upper line in Figure 5-9 c indicates a small decrease in energy use per litre oil at a preheating temperature of 140°C. which was during dual pass.2% 84 84 103 Energy input (kJ/l) - 39 36 86 54 63 69 - 5. The temperatures indicated for the Sayari expeller are the highest temperatures that occurred. For normal filtering 35 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Table 5-2 General test results for additional experiments on seed pre-treatment.4% 91.0% 88. Furthermore the variation is within the error margin and therefore uncertain. Residual oil BT50 normal crushed water 70°C water 80°C Sayari normal crushed water 80°C 11.2% 77.5% 86.9% 6. The results show that the assumptions do not hold for Jatropha as the energy reduction is negligible and the oil recovery is lower at pressures comparable to the standard situation. Boiling showed a very high oil recovery after single pass in the Sayari press of 88. which shows a small increase in the amount of energy required for pressing seed that were pre-heated at higher temperatures while processing time is constant.8 9. The BT50 data in Table 5-2 suggest that the oil-point pressure is lower for the cooked samples as their oil recovery is highest while pressure are significantly lower than those for dry seeds. Power 900 800 700 Energy input per liter crude oil 900 800 700 600 500 400 300 200 100 0 Energy input (kJ/kg) 600 500 400 300 200 100 0 100 120 140 160 180 Seed temperature (degrees Celsius) Figure 5-9 c Influence of seed input temperature on energy requirement for the BT50 press expressed per kg and per litre crude oil 5.4 142 139 312 279 6. The results are presented in Table 5-2.7% 98 92 120 141 85 82 100 115 122 120 220 198 9. Cracking was expected to decrease energy use and reduce the oil-point pressure (Faborode and Favier. Oil recovery values for Sayari expeller are after dual pass. After one week settling time approximately 30 to 50% of the volume consists of sedimentary deposits. 1996.7 Influence of other pre-treatments In addition to the previously mentioned variables some other test were conducted although in less detail.2% 89.2 9. The cooking pre-treatment also shows a positive effect in the Sayari expeller. This indicates that the engine encounters slightly more resistance.0% 87.9% 7.

Table 5-3 Oil quality analysis performed by ASG Parameter Acid value Phosphorous content Water content Oxidation stability 110°C Carbon residue Processing oil temperature Unit [mg KOH/g] [mg/kg] [mg/kg] [h] [%] [° C] Method Rapeseed standard DIN EN 14104 2.7% moisture acid value and phosphorous content are too high.4 2. 5. At high residual oil content of around 12-15% repressing of the press cake is possible. Acid value tends to increase with moisture content. More details on the presses used for testing are given in Appendix L together with a table showing industrial oil recovery rates for oilseeds.0 DIN EN 14107 15 DIN EN ISO 12937 750 DIN EN 14112 5.3% moisture 3.9 Oil quality parameters Oil quality was judged base don the five parameters shown in Table 5-3. Three very diverse samples of filtered oil were analyzed by ASG (Analytik-Service Gesellschaft mbH. The Sayari expeller is less suited for experimental use but operates much better for full time large capacity production. which are assumed to be the parameters that affect oil quality most strongly.7%. These samples cover both the widest range in oil temperature and moisture content.1 25.2 2. Germany). Oxidation stability is increased by increased water content.6 10.3% is probably due to lower temperature. Testing results are shown in Table 5-4.8 16. The other values are in range.14% phosphorous content reduces close to 15 as required by the standard at the cost of acid value.10 Comparison of suitability for Jatropha processing between BT50 and Sayari Based on the results presented in this chapter higher oil recovery is achieved in the Sayari expeller. At decreased moisture level of 2.2 0.4 21. At 13. For research purposes the BT50 is preferable because of its small capacity.14% moisture level. KG .Chapter 5: Results and discussion techniques only 5-10% sediment is acceptable (Ferchau. Some possible explanations are given here. The BT50 reaches higher oil recovery after a single pass. Rienartz additionally reported seed input temperature of 20°C and oil temperature between 30 and 60 °C increasing from feed section towards the 36 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .2 23.7%-13. Neuss. The higher carbon residue for 13.7% and then decrease form 6. Remscheid. When properly operated jamming of the press can be avoided whereas this is not possible in most cylinder-hole presses like the BT50. Batches of around two kg are sufficient for stable operation and relatively easy to condition. None of the samples comply with the PPO norm for rapeseed in Appendix B.4 97 85 52 5.3 0.4 741 733 1622 12.5 3. which is also the amount of sediments present in most other plant oils after mechanical extraction.2 0.0 DIN EN ISO 10370 0. It was not possible to determine if and how pre-treatments affect the amount of sediments. At 2. continuous processing is not possible in the BT50 due to overheating and jamming.14% moisture 6. For 6. resulting in high oil recovery. Germany) and Reinartz (Maschinenfabrik Reinartz GmbH & Co.3% moisture might be related to the darker colour of the oil due to dissolved pigments. However due to the lower residual oil content and the different press geometry dual pass is not possible for Jatropha in the BT50.3% moisture acid value and water content are unacceptable and other values are within limits. Neusäss. the point at which oil recovery was highest.3% or vice versa. 2000). The strong decrease in phosphorous content at 13.1 Test conducted by industrial press manufacturers To gain additional knowledge on the pressing process the two German press manufacturers Keller (KEK EGON Keller GMBH & Co. Severe jamming problems were however experienced when operating at restriction sizes below 9mm and moisture levels below the natural 6.10. 5. Germany) were contacted for press tests with Jatropha seeds.7% moisture 13. The data in Table 5-3 does not provide proper trends as most parameters decrease between 2.14-6. The table in Appendix K shows more detailed values of the amount of foot for various samples.

Depending on the desired oil recovery pre-treatments might therefore not be necessary. A pre-treatment of one hour cooking followed by seed drying results in the highest oil recovery for both presses. The advantage of the Keller and Reinartz presses compared to the Sayari expeller is that they require only single pass of the material to reach a high oil recovery. The influence of seed pre-heating and flaking/crushing proved to be insignificant for Jatropha seeds. Restriction size and rotational speed of the screw are two other influential parameters. Unfortunately the oil yield using this setup could not be determined due to unexpected breakdown of the press in the final testing stages. After dual pass oil recovery for the Sayari increased to 91%.Chapter 5: Results and discussion cake discharge.11 Conclusions The most important findings based on the practical research conducted for this study is presented here.) appears to be close to optimal. Seed moisture content shows the strongest effect on oil recovery from Jatropha.3-91. effect on oil recovery. The natural moisture content of the seeds (6. This pre-treatment can thus be used to decrease energy consumption and increase processing capacity. For comparison normal seeds yields were 79% for the BT50 (single pass) and 88% for the Sayari (dual pass). The energy requirement is of subordinate importance as for the BT50 the energy input per litre oil varies between 1% and 2. being close to the theoretic maximum of 95%. which is mainly attributed to higher residence time.5% of the calorific value of Jatropha oil.b. Lower rotational speeds lead to higher oil recovery. Taking into consideration all the test results optimal oil recovery is expected at a moisture content of 2-4% after cooking at 70°C.7% 5-6. Decrease in hull content below a threshold value between 66-80% has a strong. To summarize an overview of the effects of pressing parameters on the efficiency and conditions inside the press barrel is visible in Table 5-5. Restriction size should be reduced as far as possible taking into consideration both operational limits of the press and the effect of temperature on oil quality.5% oil yield [oil/oil]* 72% 79% 89% 88. The choice of rotational speed varies with the demands of the user. Single pass pressing of cooked seeds in the Sayari expeller thus leads to higher yields than normal seeds after dual pass pressing.7% 6.1% 3/7/2007 standard settings 5.3% 6.7% 14. For high production capacity rotational speed should be high and for high oil recovery per kg input material it should be low. 100% hull content and the smallest restriction size and lowest speed possible for a certain press type.0% 6.7% 11.% w.0% 6. Single pass pressing of cooked and dried seeds resulted in the highest measured oil recovery of 89% for both BT50 and Sayari press. Oil recovery of these machines is higher than for the BT50 and similar to the Sayari press. although negative.7 wt. * [% oil/oil] means the percentage of the oil present in the seeds that has been obtained by pressing KEK Keller p0101 1 2 3 Reinartz AP08 date 4/6/2007 4/6/2007 6/7/2007 sample description standard settings start sample standard settings end sample changed pressure cones moisture oil contentpress cake 6. 37 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Speed can only be reduced to a limited extent as below 20 RPM operation is not possible. Table 5-4 Results from tests at KEK Keller and Reinartz conducted in 2007.

throughput and energy use Oil recovery Pressure Temperature Throughput Energy/liter Press parameters RPM restriction size Seed treatments heating flaking moisture content hull fraction boiling - - - - 38 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .Chapter 5: Results and discussion Table 5-5 Overview of the effects of press settings and seed pre-treatment on oil recovery. temperature. pressure.

2 shortly explains CBA theory. Although one can distinguish between a financial. Paragraph 6. Some of the experimental results were used to make recommendations on how to set up Jatropha oil production in rural Tanzania. Data on the costs and benefits of small scale Jatropha oil processing for rural communities was gathered during a visit in Tanzania from January to April 2007.3. Is Jatropha just another opportunistic outburst trying to find a solution to global poverty problems or is it really possible to fight poverty using Jatropha? Despite all euphoria concerning Jatropha knowledge on how to actually process it into oil as to make it profitable for rural communities is fairly limited. Prices for transport and labour are estimates based on Tanzanian prices. Subsequently paragraph 6. 6. 2005). economic. 6. Prices of machinery were obtained from the German screw press manufacturers Reinartz and Keller.Chapter 6: Possibilities for oil pressing in rural Tanzania 6 Possibilities for oil pressing in rural Tanzania This chapter puts the results on Jatropha expression from previous chapters into a broader context.2 Cost Benefit Analysis: Theory The CBA is designed as a tool to help investors to select projects that will contribute most to their objectives while keeping in mind their limited financial resources (Romijn and Biemond. taking questionnaires and the writers interpretation of the situation at hand. Goal To see whether Jatropha has potential to improve the situation in poor tropical countries one should think of how to setup local production.1 Introduction Occasion Over the last decades the western approach of setting up development projects in less developed countries has proven extremely ineffective (Douthwaite. Conclusions are drawn in paragraph 6. Most of the data was obtained by interviewing. 39 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . This method has been developed by the World Bank and the United Nations Industrial Development Organisation (UNIDO) and is often referred to as project appraisal. social and environmental CBA this report will only focus on the financial aspect as this is relevant for the question at hand and fits well to the available data. Research outcomes are presented and discussed in paragraph 6. The main goal of this chapter is therefore to find out whether it is preferable to centralize or decentralize Jatropha oil production when decentralized cultivation of Jatropha is assumed? Furthermore calculations are done to determine if Jatropha can be more than just a primary agricultural product in rural areas in less developed countries Approach For local communities to financially benefit from Jatropha cultivation probably requires cooperation between investors and small farmers. However economic viability in itself is not sufficient to establish successful projects thus some local circumstances are taken into consideration before making recommendations.5. Centralized and decentralized production of Jatropha oil are compared using the cost benefit analysis (CBA) method. As an extension another CBA was conducted to judge the viability of oil production on village level. 2002). Structure The setup of the rest of this chapter is as follows.4 passes judgement on the viability of Jatropha oil production on village level.

For IRR =< market interest rate ‘i’ the yield is equal to or less than the cost of financing.org/publications/inflationDevelopments. Sensitivity analysis 6. 2005). visited 04-06-2007 2 Source: Website of the Bank Of Tanzania: http://www.9%2 1 Source: website of the Bank Of Tanzania: http://www. Depreciation is not taken into account in a CBA as it leads to costs without leading to expenditures.2. For IRR > market interest rate ‘i’ the project is more profitable than the cost of financing the project. The core of the CBA analysis is to compare the earnings profile of the project to the going interest rate which represents the costs of financing as discussed above (Romijn and Biemond. This is best explained looking at depreciation. IRR The IRR is a measure for the extent of profitability of the project compared to the going interest rate and is determined by setting the NPV to zero. High inflation rates make it difficult to estimate prices of inputs 5 or 10 years from now.bot-tz. The effects of inflation and interest are then combined in the discount factor: Discount factor = with 1 (1 + r ) t t = number of years r = real discount rate r= (1 + i ) −1 (1 + p ) with i (interest rate) = 15. Multiplying non-financial cash flows by the discount factor yields discounted cash flows. and between income and revenues is not relevant. Profitability assessment 2. The same holds for benefits and income. A NPV =< 0 indicates no commercial reason to embark in the project. For borrowed money the interest rate on loans is applied while for own money the interest rate on savings is used. 2005). Constant prices are the prices in the base year of the CBA. Inflation Because of Tanzania’s significant inflation rate it is convenient to work with so called constant prices. NPV The NPV is the sum of the discounted non-financial cash flows during the projects lifetime. For an NPV > 0 a project is expected to yield more than the interest rate. Liquidity assessment and determination of a financial plan 3. Both criteria use discounting techniques which take into account the effect of interest that would be received if the money was in the bank.org/publications/EconomicIndicators/ InterestRates. As a CBA takes the entire life-time of a project as its time horizon total costs will be equal to total expenditures by definition (Romijn and Biemond. The most frequently used criteria in this comparison are the Net Present Value (NPV) and the Internal Rate of Return (IRR). visited 04-06-2007 40 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .54%1 p (rate of inflation) = 6.bot-tz. Normally a CBA consists of the following three steps: 1.htm. The advantage of using constant prices is that the same prices can be used without taking inflation into account.1 Profitability assessment When judging project profitability the way of financing determines the interest rate to be used to judge the investments profitability. Therefore the project is viable in principle although liquidity and risk assessment should be conducted.Chapter 6: Possibilities for oil pressing in rural Tanzania In a CBA the distinction between cost and expenditures. thus providing the exact sum of money which a project is expected to generate.htm.

labour costs. As liquidity does not add anything to the question asked in this chapter it will not be taken into further consideration. Sediment level is 40 vol% and contains 55 wt.Chapter 6: Possibilities for oil pressing in rural Tanzania 6. Prices are all given in dollars as this is the foreign monetary unit used in Tanzania. Lifetime of German presses is 20 years. The CBA was used to compare centralized and decentralized processing of Jatropha seeds into oil and press cake.3 Sensitivity analysis As projects are always subjected to uncertainty one should assess the sensitivity of the results of the CBA to changes in specific costs and benefits. 6. Instead the number of trucks needed changes with travel distance and production capacity. 2005). Supply of seeds is not taken into account in this respect. The selling price for Jatopha oil is 750 TZS/l which is equivalent to $0. Obviously the distance along which input material has to be transported is considerable reduced in case of decentralized processing. assumptions and calculations.2. By calculating so called switching values (the degree of change in a value that sets NPV to zero) one can determine the risk of a project. For increased production capacity the amount of kilometres a truck cover per year remain the same.000 based on a small Mercedes truck and reducing the price to Tanzanian levels. F and K.3 CBA results: comparison of centralized and decentralized production This paragraph focuses on the most viable production options for centralized and decentralized production. This is done by executing a sensitivity analyses in which only one variable at a time can be changed to see how this affects NPV and IRR. equipment maintenance costs and revenues. The difference between central and dispersed positioning of pressing facilities is represented in Figure 6-1. In CBA calculations write-off in 5 years is assumed for the Sayari. In addition to difference in press type and size different oil cleaning methods were compared. The main variables included in the CBA are investment cost for equipment. while Tanzania Sayari expeller needs to be replaced after 5-10 years.2 Liquidity assessment In principle the liquidity assessment determines whether a project is able to fulfil payment obligations from its cash inflows during its entire lifetime (Romijn and Biemond. The main assumptions are listed below. More details on the different presses is given in Appendix C. As can be seen in the table a comparison is made between three different processing configurations. The values in Table 6-1 were computed with a combination of data sources. This price is reasonable taking into account that fuel taxes have not yet been included in the calculation. • • • • Cost price of an 8 ton capacity truck is $15.% oil. Main assumptions • From previous chapters the oil recovery is set at 85%. 6.2. 41 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . The cases for centralized pressing incorporate only one location and one machine. In case of decentralized pressing the values represent expenses for all machines at multiple locations. Table 6-1 shows the input values used for calculations. Production capacity is a measure for the oil production.59. seed purchasing price. The results for different types of presses and filtering methods can be found in Appendix M.

3. The 5 dotted circles indicate the areas for decentralized production with their own pressing facility. One operator is required for each press independent of its size. Two seed collectors are assumed to be on each truck. Figure 6-1 Area division for centralized and decentralized production. 42 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . centralized and large scale centralized 24h/day production respectively. The large circle represents centralized pressing which is located at the hatched circle.Chapter 6: Possibilities for oil pressing in rural Tanzania Therefore write-off is 10% in all cases.05/kg without further processing. • • • Press cake is sold for $0. 6 and 12 trucks are assumed for decentralized.

316 5.org www. The values in Table 6-2 to Table 6-8 are best guesses.200 $153.299 $3.00 $0.062 700 Reinartz & KEK Egon Keller 1.271 $261.08% 5 x KEK P0101 (decentralized) 1 x Reinartz AP14/30 (centralized) Reference www.675 6 $0.Chapter 6: Possibilities for oil pressing in rural Tanzania Table 6-1 Input values for the cost benefit analysis Monetary exchange rate (dollar to chilling) discount factor interest rate inflation rate real discount rate Press type 1268 20.214 $15.411 Reinartz 14000L $32.713 Diligent Tanzania LTD $41.655 $3.877 $121.178 $30.31% 15.280 Reinartz & KEK Egon Keller 47.850 1.606 Reinartz Maintenance annual maintenance (5%) capacity (kg Jatropha/hr) annual capacity ton (8 h operation) annual capacity ton (24 h operation) Seed buying costs price/kg 8h 24h Fuel use for pressing Annual fuel use 8h (liter) Annual fuel use 24h (liter) Amount usable oil 8h (liter) 24h (liter) Fuel costs capacity (ton/truck) 8h number of rides 8h average km/ride total km 8h km/liter price /liter total fuel 8h total fuel 24h truck price $6.00 $0.08% shows this option is quite profitable.09 $62.bot-tz.1 Decentralized pressing Decentralized processing in all cases showed negative NPV at 8 hour operation per day as can be seen in Appendix M.org www.704 5000L $115.038.000 8 Mercedes 815 Atego 165 750 123.527 $0.bot-tz.99 $10.3.726 $123.914 8000L $26.bot-tz.539 $0.200 $119.00 $0.054 $1.018 $23.677 $23.071 $15.400 39.09 Diligent Tanzania LTD $124.576 $519.356 350 658 2.org www.165 861.535 $15.726 $247.688 11.376 Reinartz & KEK Egon Keller 205.000 8 82 750 61.200 $109.764 $4.047 $21. Centralized processing was possible at 8 hour operation although 24 hours was also strongly preferred.271 $261.965 $232.452 6.030 $96.893 $41.538 $34.229 $5.754.45 $842 Diligent Tanzania LTD $22.486 417.100 $48.304 $101.125 $101.90% 8.060 $193.069 $82.352 $509.243 Reinartz & KEK Egon Keller $140.000 Mercedes 815 Atego Labour cost wage dollar/month wage /hr 8h press operator seed collectors manager total 24h press operator seed collectors manager total Revenues oil 8h oil 24h presscake 8h prescake 24h $80.348 350 658 2.09 $62.694 208.356 $41.107 $80. The positive value for equipment in year 5 represents the NPV of the equipment taking into consideration the difference in write of period between German and Tanzanian 43 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .539 $0. Table 6-2 shows a positive NPV for decentralized processing during 24 hours a day.035 $46.678 $82.bot-tz.54% 6.139 Diligent Tanzania LTD $82.99 $20.543 $523.99 $4. When comparing the real IRR of 34% to the real interest rate of 8. Rounded numbers would only increase inaccuracy.808 $80.538 Diligent Tanzania LTD $68.071 $12.640 23.079 9.480 5.030 $96.375 6 Mercedes 815 Atego $0.688 6 $0.100 $64.org 1 x Reinartz AP15/45 Data Source (centralized) Investment press cyclone filter storage $126.972 8 82 300 24.765 $87.357 $61.45 $842 $11.764 $5.45 $5.151 $1.100 $53.925 877.

163 0.539 -$12.892 NPV $605.132 0.000 -$310.247 $615.315 -$261.696 $606.482 -$261.539 -$30.539 -$13.071 -$153. 10 and 20 years are taken for respectively Tanzanian machines.535 -$119.808 -$7.436 -$105.090 $195. Table 6-3 shows a significant increase in both NPV and IRR compared to using the more expensive German technology.696 $606.877 -$10.337 0.977 1.124 $271.696 $606. As can be seen from the IRR of 61% in Table 6-4 profitability of the project increases significantly when production is centralized compared to the situation in Table 6-2.915 -$261.185 IRR 61% Judging from Table 6-5 economies of scale seem to have more effect on profitability than the way production is organized.808 -$7.231.856 $182.877 -$10.403 $213.000 -$567.569 -$12.744 $196.163 0.856 $183. Comparing Table 6-4 and Table 6-5 five years NPV increased with over $1.214 -$105.744 $392.994 IRR 99% 44 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .569 -$12.214 -$105.569 -$12.977 -$261.925 $181.403 $213.877 -$10.124 -$523.539 -$30.079 -$61.384 0.384 0.487 $1.000 for a $170.539 -$13.214 -$105.079 -$61.436 -$105.575 -$329. Table 6-2 Calculation of NPV and IRR for decentralized oil production capacity of 350kg/hr 5 x KEK P0101 (70kg/hr) 24 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$567. Write-off periods of 5.231.528 -$261.000 -$481.124 Revenues $1.840 -$261.240 -$523.436 -$105.925 $197.808 -$7.000 increase in initial investment.403 $214.487 $1.247 $615.792 $168.071 -$153.733 $157.539 -$12.696 $606.686 -$523.247 $615.678 $401.535 -$119.403 $214.000.203 $483.569 -$12.808 -$7.337 0.614 0.744 $196.856 $168.696 $606. Table 6-3 Calculation of NPV and IRR for decentralized oil production capacity of 350kg/hr using the cheap Sayari expeller 5 x Sayari (70kg/hr) 24 hours operation per day with cyclone filter and cheaper storage Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$329.539 -$12.575 -$261.578 NPV $538.792 $169.535 -$119.877 -$10.384 0.539 -$13.696 $606.569 -$12.733 $144.539 -$12.487 Total Discount factor Dicounted Cash flow -$481.518 -$261.569 -$12.678 $511.337 0.614 0.436 -$105.837 $483.076 -$261.627.760 1.686 $483.792 $382.925 $447.614 0.029 -$261.539 -$30.569 -$12.403 $411.808 -$7. Table 6-5 Calculation of NPV and IRR for centralized oil production capacity of 750 kg/hr 1 x Reinartz AP15/45 (750kg/hr) 24 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance -$481.384 $197.216 $754.071 -$153.113 -$261.744 $196.337 0.877 -$10.733 $354.807 0.214 -$105.2 Centralized pressing Centralized production at equal production capacity results in a higher NPV due to reduced investment costs.678 $278.000 -$329.760 -$261.247 $615.238 0.079 -$61.732 -$261.654 NPV $1.124 -$523.231.539 -$12.413 0.487 $1.218 $379.403 $214. At doubled production the IRR increases from 61% to 99% making the much more profitable.569 -$12.079 -$61.696 $606.403 $214.Chapter 6: Possibilities for oil pressing in rural Tanzania machines.856 $413.403 $592.231.569 -$12.792 $155.231.539 -$13.071 -$153.227 -$261.196 IRR 63% 6.535 -$119.744 $196.079 -$61.966 NPV $658.760 -$310.614 0. Table 6-4 Calculation of NPV and IRR for centralized oil production capacity of 350 kg/hr 1 x Reinartz AP 14/30 (350kg/hr) 24 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$310.163 0.977 -$567.487 $1.367 IRR 34% Investment cost can be reduced by selecting local technologies like the Sayari expeller.539 -$13.925 $198.575 1.163 0.770 $483.403 $213.071 -$153.733 $156.577 0.403 $213.696 $606.247 $615.696 $606.436 -$105.124 -$523. trucks and German machines.678 $265.535 -$119.3.569 -$12.357 -$261.686 1.539 -$30.696 $606.214 -$105.539 -$30.

000 -33% 54% $830.000 +497% -4% -$50.000 -27% Table 6-7 Sensitivity analysis centralized production Sensitivity analyses 24h operation Reinartz AP14/30 Cash flow item Equipment costs Seed costs Fuel costs Labour costs Revenues Jatropha oil price Assume 5% increase 25% increase 10% increase 10% increase 10% decrease 20% increase Effect IRR Effect NPV Switching value -3% -$15.000 +669% -3% -$60.000 +58% -1% -$12.000 +129% 10% decrease -11% -$240. Changes in raw material price and oil selling price and revenues show the biggest impact. In addition Table 6-8 shows sensitivity decreases for larger production scale. Table L-3 in Appendix L shows that the differences are expected to be within 1-2% for most oilseeds When production is centralized large presses should be used to reduce investment costs. For 24hr/day operation the project shows little risk for both centralized and decentralized production according to the sensitivity analysis and can therefore be implemented in any of the three setups. Table 6-6 Sensitivity analysis decentralized production Sensitivity analyses 24h operation KEK P0101 Cash flow item Equipment costs Seed costs Fuel costs Labour costs Revenues Jatropha oil price Assume Effect IRR Effect NPV Switching value 5% increase -2% -$30.000 +1116% 10% increase -2% $40. Normally transport costs are expected to contribute to a larger extent.000 -29% Table 6-8 Sensitivity analysis centralized production Sensitivity analyses 24h operation Reinartz AP15/45 Cash flow item Equipment costs Seed costs Fuel costs Labour costs Revenues Jatropha oil price Assume 5% increase 25% increase 10% increase 10% increase 10% decrease 20% increase Effect IRR Effect NPV Switching value -5% -$25.000 +78% -1% -$25.000 +52% 10% increase 0% -$4.000 +127% -21% -$245. which is to be expected.000 +266% -26% -$490.3. Revenues have the lowest switching variable which still has a margin of >20% in all cases.000 -39% 6.4 Non financial consideration Determination of the optimal pressing arrangement should not solely be based on financial considerations.000 -25% +34% 400.000 + 95% 25% increase -12% -$260.000 -23% 20% increase +29% $400.000 +338% -26% -$520.Chapter 6: Possibilities for oil pressing in rural Tanzania 6. Some additional factors that should be taken into account during decision making are listed here: • The oil recovery for large presses is only slightly higher than for smaller presses because the design is often very similar.3 Sensitivity analysis A sensitivity analysis (see table 6-6 to 6-8) of the different cases shows centralized production is less sensitive to fluctuating values than decentralized production. Local people in less developed countries are generally unfamiliar with these • 45 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Clearly the larger the production scale the less sensitive the process is to the changes taken into account in this paragraph. The extremely high switching value for fuel costs in all three cases suggests underestimation of travel distances and fuel consumption.3.000 +195% -22% -$260. The project shows very little sensitivity to changes in most variables.

This could in turn lead to improved product quality and yield as pressing locations could function as knowledge centres. Although locals are perhaps unable to technologically modify the screw press they are able to come up with new applications. but one should take care not to confuse our needs with the needs of the local population. An example is the use of a multi functional platform for pressing. The role of product champion could for example be filled by a village leader. The innovation champion will also promote the beneficial applications of the new technology. • • • This short consideration of innovation theory was restricted to some of the significant success factors. As is the case for the village Temi ya Simba there is no grid connection present in the village. There might of course be a need for the advantages of using Jatopha for electricity or cooking. • • Incorporating some points mentioned in the ‘learning selection model’ from Douthwaite (Douthwaite. Investments costs are relatively low and capacity can be gradually increased. The fact that most trees were not harvested in areas where Diligent was actively buying suggest lack of motivation on the side of the farmer. milling and electricity generation. Linking Jatropha to such directly noticeable benefits can increase motivation in local communities. For decentralized pressing 80% of the production capacity remains in case of a single breakdown and repair cost will be lower. This might be caused by low prices or the inability of local farmers to judge the long term benefits. • Small decentralized presses allow more flexibility in the expansion of production capacity. The last point mentioned here is the necessity of a product champion. In case of centralized pressing operational mistakes could cause the whole production to come to a halt or cause expensive machine damage. This is someone who is respected in his community and who works with users and producers to identify and solve problems in order to increase the fitness of a technology. Most remote villages have no direct application for Jatropha oil and to them Jatropha is just another cash crop. In case of problems Tanzanian people prefer personal visits over telephone or email. With respect to Jatropha the activities of a product champion could be to motivate farmers to grow Jatropha and to make them aware of long term benefits of increased income security. Therefore a Multi Functional Platform (MFP) was installed by a 46 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Scattered production locations reduce the distance between farmer and oil producer and can therefore improve communication between both parties. This issue is questionable fro Jatropha oil. Expansion into new geographic areas is easier with decentralized production as fewer investments are required to setup new pressing units at other locations. 6.Chapter 6: Possibilities for oil pressing in rural Tanzania large and complicated machines.4 Viability of community level processing: Engaruka case The village of Engaruka consists of two separate parts: northern and southern Engaruka. Most of the 9000 inhabitants of this Maasai village make a living as farmers or keeping feedstock. 2002) gives rise to some additional comments: • Judging from experiences during field trips in Tanzania local farmers seemed moderately motivated to grow Jatropha or collect seeds from wild Jatropha trees. However the innovation theory discussed by Douthwaite contains many other factors that should be satisfied. According to Douthwaite the need of an innovation should be great.

Most of the villagers cannot get any loans from the bank as farming is excluded from the definition of enterprises the bank uses when judging loan requests. The total land area available for Jatropha is around 100 acres. Figure 6-2 Jatropha utilization in Engaruka The possibility of village level processing of Jatropha seeds into oil was evaluated using a method similar to the previous paragraph. As Jatropha was already grown in Engaruka for fencing most seeds still stem from hedges. Figure 6-2 shows a model for Jatropha utilization in Engaruka. In addition to the grain mill connected to the MFP seven more diesel powered grain mills are present in the village. 2005. In case of local processing Jatropha oil is sold at $0.Chapter 6: Possibilities for oil pressing in rural Tanzania Tanzanian NGO named TaTedo in November 2006 to combine seed pressing with electricity generation and grain milling. The first harvest in 2007 produced approximately 2400 kg Jatropha seeds. Over 30 farmers are growing Jatropha on small plots between 0. These savings or shares form the basis for providing credit to members. A press cooperation consisting of maybe 5-10 people could start production of Jatropha oil and grain millers could also benefit from this cheaper locally produced fuel to power their engines. The importance of this market segment is underlined by a SACCOs market of nearly US$1 billion in Kenya. Together with the diesel driven generator these machines constitute the potential Jatropha consumption in Engaruka. 193-202).59 and press cake is used as fertilizer. 3 houses and a campsite are provided with energy from this platform. In total 20 shops. Therefore Savings and Credit Co-operatives (SACCOs) are included in this system. The available credit is normally around three times the level of savings or shares. The MFP operates five hours a day generating roughly 6 kWh. 47 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .5-3 acres in size (Zheng and others. which involves pooling of voluntary savings from members in the form of shares. The investment required to setup a press-cooperation probably exceed the amount of money available to the local investors. Assumptions in this comparison are: • • 100 acres is available for growing Jatropha without decreasing food production. Calculations shown in Appendix N show that although village level pressing is profitable larger profits are achieved by just selling seeds. The SACCO system is a mutual membership organization.

This method was used to select the best option from a range of possibilities. However.500 - Judging from Table 6-9 pressing on village level can contribute to the welfare level in Engaruka. More information on all three villages can be found in Appendix O. For similar production scale centralization allows an IRR of 61% compared to 30-40% in case of decentralization. The availability of oil and electricity are of additional value to the whole village.Chapter 6: Possibilities for oil pressing in rural Tanzania • When only selling seeds investment costs are assumed to be zero and fertilizer costs are equal to the value of the fertilizer produced in case of processing seeds to oil ($0. need of a technology and the presence of a product champion. One of the villages only sells seeds and in the other limited quantities of oil could be pressed to fuel cooking stoves. The initial investment costs for Jatropha oil production are high compared to expected revenues. When production is running 24 hours a day. Because of its small scale village level production of Jatropha for cooking is expected to be inefficient taking into account the modest profits in the much larger Engaruka case. Jatropha has significant added value for rural communities in Tanzania. 6. For good yields the additional income can amount up to one hundred annual wages.5 Conclusions One should keep in mind that the CBA explained in this chapter provides insight in the profitability of a project.750 and $54. especially when taking into account the increase in welfare due to availability of oil and electricity. For a village of 9000 inhabitants and 100 acres of available land the NPV of Jatropha cultivation appears to be between $12. which is 48 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .950 $14. The CBA for Engaruka suggests that Jatropha activities on a scale smaller than this should be restricted to direct selling the primary product. As was the case when comparing centralized and decentralized processing in the previous paragraphs larger production means higher returns (see Table 6-9). Labour costs are zero as farmers grow Jatropha for additional income besides their normal activities.500 per year depending on seed yields. Jatropha use in these villages was not as extensive as in Engaruka. Daily production time of 8 hours is expected not to be profitable based on the executed CBA. 7 days a week production showed profitable. In case of 100 acres and low yield the increased revenues are equivalent to 19 monthly salaries. In case of noncommercial projects the use of local technology is desirable as this creates financial activity and potentially yields higher revenues.05 /kg). Calculations for selling seed include no local electricity and oil production. Commercial firms might prefer more robust technology.500 33% 115% 353% 100 acres low yield seeds $12. reviewing both financial and non-financial aspects shifts preference to decentralized pressing for Tanzania and similar countries. which is much more reliable and durable at the cost of higher initial investments. Based on purely financial results from the CBA large scale centralized processing is preferred over decentralized pressing.500 $51. produce electricity and mill maize. Initially two other villages were reviewed on Jartopha application.750 100 acres good yield seeds $54. Calculations for Jatropha pressing assume a small village cooperation that uses an MFP to press seeds. Calculated NPVs are no indication for the absolute value of commercial profits. • • • Table 6-9 NPV and IRR for various situations in Engaruka NPV IRR 50 acres low yield 100 acres low yield 100 acres good yield $2. For projects to become successful care should be taken to bear in mind the importance of local circumstances such as motivation.

500 per annum. Oil processing for 100 acres of Jatropha could easily provide an additional $1.Chapter 6: Possibilities for oil pressing in rural Tanzania equivalent to approximately 100 annual minimum wages. With this extra income comes the availability of electricity and cheap fuel. In depth investigation of financing possibilities through micro-credit is advisable. 49 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Oil processing on village scale without the intervention of commercial firms seems preferable to mere seed selling if coupled to electricity generation and maize milling. which is equivalent to 19 monthly salaries to be divided between the initiators.

Depending on the desired oil recovery pre-treatments might therefore not be necessary.5% of the calorific value of Jatropha oil. For high production capacity rotational speed should be high and for high oil recovery per kg input material it should be low. For comparison normal seeds yields were 79% for the BT50 (single pass) and 88% for the Sayari (dual pass). from 80-100% the yield is hardly influenced. Speed can only be reduced to a limited extent as below 20 RPM operation is not possible with the BT50. being close to the theoretic maximum of 95%. Single pass pressing of cooked seeds in the Sayari expeller thus leads to higher yields than normal seeds after dual pass pressing. although two different press types were included in this study. Only the first two could be included in the present study. Lower rotational speeds lead to higher oil recovery. For all tests the amount of foots (30-50 vol. • • What is the optimal setup for the production of Jatropha oil in Tanzania? In what way can rural areas in which their use is anticipated benefit from Jatropha cultivation? 50 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . However standard treatments are likely to bring the quality parameters within the range of these standards. The influence of seed pre-heating and flaking/crushing proved to be insignificant for Jatropha seeds. 100% hull content and the smallest restriction size and lowest speed possible for a certain press type.) appears to be close to optimal with an oil yield of 79%. If high yield are required restriction size should be reduced as far as possible taking into consideration both operational limits of the press and the effect of temperature on oil quality. Both seed pre-treatment and press operation showed significant impact in oil recovery and energy requirement. the latter of which can be due to both seed pre-treatment and press operation.Chapter 7: Conclusions and recommendations 7 Conclusions and recommendations The main question to be answered by this study is as follows: • How can oil recovery. changes in press operation and changes in press design. This pre-treatment can thus be used to decrease energy consumption for pressing and increase processing capacity. Oil yield for different moisture levels varies from 3388%. Single pass pressing of cooked and dried seeds resulted in the highest measured oil recovery of 89% for both BT50 and Sayari press. Restriction size and rotational speed of the screw are two other influential parameters. The energy requirement is of subordinate importance as for the BT50 the energy input per litre oil varies between 1% and 2.7 wt. None of the samples tested for quality met standards for Pure Plant Oil.b. A pre-treatment of one hour cooking followed by seed drying results in the highest oil recovery for both presses. Seed moisture content (varied from 2-13%) shows the strongest effect on oil recovery from Jatropha. Quality was mostly affected by moisture content and temperature. Taking into consideration all the test results optimal oil recovery is expected at a moisture content of 2-4% after cooking at 70°C. energy input and cost of Jatropha seed expression in screw presses be optimized while retaining a quality that complies with international fuel standards? Three main options to increase oil recovery have been identified: pre-treatment of seeds before pressing. Oil yields drastically decrease below 80% hull content. which is mainly attributed to longer residence time.% w. The choice of rotational speed varies with the demands of the user. Additional consideration of the practical application of screw presses for the production of Jatropha oil in Tanzania led to answers to the following two additional questions. The natural moisture content of the seeds (6.%) in the crude oil was problematic for filtering. After dual pass oil recovery for the Sayari increased to 91%.

reviewing both financial and non-financial aspects shifts preference to decentralized pressing for Tanzania and similar countries. Repetition of press tests and measurements on oil recovery would increase the reliability of the results. Some recommendations for further research are listed below: • This study was only related to screw pressing process. When production is running 24 hours a day. 7 days a week production showed profitable. A closer look at the success factor given by Douthwaite is advisable when actually starting a project for Jatropha processing in any developing country. which is only part of the production process of Jatropha oil. central and high capacity central). In case of non-commercial projects the use of local technology is desirable as this creates financial activity and potentially yields higher revenues. The main reason is higher flexibility to adapt to various local circumstances. Optimization of industrial presses for Jatropha seeds might still be interesting to recover a few additional percentages of oil and reduce the amount of foot in the crude oil. Commercial firms might prefer more robust technology. Additional research aimed at designing the total production process would provide better insight on process efficiencies and viability. Solutions for filtering the crude oil could for example result in higher yields of clean oil than optimizing the oil recovery yield of a screw press. which is much more reliable and durable at the cost of higher initial investments. which is equivalent to 19 monthly salaries to be divided between the initiators. In depth investigation of financing possibilities through micro-credit is advisable with a view to planning community Jatropha projects in Tanzania. The CBA method was used to select the best option from a range of possibilities (decentralized. which is equivalent to approximately 100 annual minimum wages. Jatropha appears to have significant added value for rural communities in Tanzania.500 per annum. Based on purely financial results from the CBA large scale centralized processing is preferred over decentralized pressing.750 and $54. Calculated ‘Net Present Values’ (NPVs) are no indication for the absolute value of commercial profits. Oil processing for 100 acres of Jatropha could easily provide an additional $1. For a village of 9000 inhabitants and 100 acres of available land the NPV of Jatropha cultivation appears to be between $12. Decentralized setup of Jatopha press plants is preferred over centralized processing although CBA analysis predicts higher profits in case of centralization. Oil processing on village scale without the intervention of commercial firms seems preferable to mere seed selling if coupled to electricity generation and maize milling. For similar production scale centralization allows an ‘Internal Rate of Return’ IRR of 61% compared to 30-40% in case of decentralization. Daily production time of 8 hours is expected not to be profitable based on the executed CBA. One should keep in mind that the CBA explained in this chapter provides insight in the profitability/viability of a project. With this extra income comes the availability of electricity and cheap fuel in rural areas. • • • • 51 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries .Chapter 7: Conclusions and recommendations A ‘Cost Benefit Analysis’ (CBA) was used to answer both questions.500 per year depending on seed yields. However. The initial investment costs for Jatropha oil production are high compared to expected revenues.

2007. v. Zed Booka Ltd. Jayabalan. 37. v. B. Becker. O. G. Bargale. Enabling Innovation: Apractical guide to understanding and fostering technological change: London. A Twin-screw extruder for oil extraction: direct expression of oleic sunflower seeds: Journal of the American Oil Chemists' Society. p. C. and Mouloungui. P. Broeck. Oil expression characteristics of rapeseed for a small capacity screw press: Journal of food science technology. K. N. 223-243. New Delhi. 2314-2323. Characteristics and composition of Parkia biglobbossa and Jatropha curcas oils and cakes: Bioresource Technology. Rubaza. 10-6-2006. p. J. p.Focus on Jatropha: Biodiesel: Technology and Business Opportunities ñ An Insight. p. Chemical composition and effect of heat on organic matter. Z.and nitrogen-degradability and some antinutritional components of Jatropha meal: Animal Feed Science and Technology. Technology Information. 2004. FACT FOUNDATION. 2006. H. 1999. N. and Ponraj. Evaluation and bioinduction of energy components of Jatropha curcas: Biomass and Bioenergy. v. S. p.. Dufaure. Leyris. and Srikanth. 67. 27. 76. S.. E. 130-134. Douthwaite. D. fettprodukten. 161-164. J. L. A. M. Performance and emissions characteristics of Jatropha oil (preheated and blends) in a direct injection compression ignition engine: Applied Thermal Engineering. M. 2007. V... and Agarwal. and Stein. 2006. v.. Eggers. Makkar. 1997.Reference list Reference List Jatropha Handbook. 1073-1080.. and Seiler. L. C. 52 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . 2002. Augustus.. P... K. Agarwal.. Assessment of the potential of JAtropha Curcas (biodiesel tree) for energy production and other uses in developing countries. Biodiesel Conference Towards Energy Independence .. Messung von Druck. Rigal.... 307-310. 2002. 2002. 23. Kaushik. H. S. seed collector Temi ya Simba for Diligent Tanzania in. O. and Singh.. Ref Type: Conference Proceeding DGF-Einheitsmethoden... 2000.. 92. 2006. Akintayo. R. 494-499. D. Wissenshaftliche Verlagsgesellschaft mbH. tensiden und verwandten stoffen: stuttgart. v.. Ref Type: Personal Communication Aderibigbe. Deutsche einheitsmethoden zur untersuchung van fetten.. P. Brahma Singh. G. R. A. Maj Gen.. W. Forecasting and Assessment Council (TIFAC) Department of Science and Technology.und Temperaturprofilen beim abpressen van Ölsaaten in Seiherschneckenpressen: Fette Seifen Anstrichmittel. Benge. 87. Johnson.. v. p. v. E. p. Swaminathan. G. Ref Type: Report Biswas. J. T. Ref Type: Unpublished Work Interview with Bruno S. C. and Foidl.

v. p. Rome. v. H. E. 159-160. Ref Type: Report Kandpal. Detoxification experiments with the seed oil from Jatropha curcas L: Industrial Crops and Products. 1996. Y. Influence of compound structure and comparison to petrodiesel fuel components: Fuel. Physic nut Jatropha Curcas L. O. Folkecenter for Renewable Energy. Reinhard K. 335-345. Identification and Significance of the Oil-point in Seed-oil Expression: Journal of Agricultural Engineering Research. p. 1996. 6. Ref Type: Report Faborode.. 2005. Bagani. 1995. J. The Jatropha System . p. 1999. K. M. and Steidley. 2007. Y. W. H. Ref Type: Report Heller. Mittelbach. and Trabi. 2004. v. Gatersleben/ International Plant Genetic Resources Institute. M. 1-8. R. 65. v. Khan. M. p.. L. Expression of oil from oilseeds--A review: Journal of Agricultural Engineering Research. v. (JCL) as Raw Material and as Renewable Energy. 28. Bagani. Ref Type: Personal Communication Gubitz. and Madan. K. LLC. John Wiley & Sons. Exploitation of the tropical oil seed plant Jatropha curcas L: Bioresource Technology. promoting the conservation and use of underutilized and neglected crops 1. 1059-1065. 6-4-2004. Ref Type: Serial (Book. M. M. and Mittelbach.economy and dissemination strategy. Kinematic viscosity of biodiesel fuel components and related compounds. M. Integrated Rural Development by Utilisation of Jatropha curcas L. Haas. Fracture resistance of sunflower seed and kernel to compressive loading: Journal of food engineering. Institute of Plant Genetics and Crop Plant Research. Jatropha Curcas L. Ref Type: Report Foidl.... 1996. Jatropha curcus: a renewable source of energy for meeting future energy needs: Renewable Energy. Reinhard K. Ref Type: Report Hui. 2000. 2007. A. Hui. and Favier.. and Das. S. Isopropanol injection in extruder.Monograph) Henning. 67. Knothe.Reference list Enterprising Solutions Global Consulting. 12. G. M. 2000. 53 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . B. p. 46. K. in Afrika: Assessment of the impact of the dissemination of "the Jatropha System" on the ecology of the rural area and the social and economic situation of the rural population (target group) in selected countries in Africa. Ref Type: Conference Proceeding Henning. N. UNCDF microfinance programme impact assesment 2003.. Case Study "Jatropha Curcas". M. v. 111-118. Ferchau. 2003. J. J. 495-503. 1983. 1-11-2000. Equipment for decentralized cold pressing of oil seeds. G. v. 84. R.. and Hanna.. p. p.. Gupta. F. Hartlieb Euler and David Gorriz. Bailey´s industrial oil & fat products volume 2:" Edible oil &fat products: oil and oil seeds". 73-82. Global Facilitation Unit for Underutilized Species (GFU) & Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH.

Biodiesel Conference Towards Energy Independence . p. R. and Nahar. C. Paramathma. Parthiban. Cost-Benefit Analysis of Projects. M. v. 7-2-2007. Balasubramaniam. Diligent.. 19. P. et al. Vadke. Lehada C. 7-2-2007. 1990. p.. R. T.. v. Ref Type: Personal Communication Shahidi. Mettupalayam. 2006.Focus on Jatropha: Development of Inter-specific Hybrids in Jatropha. N. 1169-1176. H. W. M. 85. 2005. Sirisomboom. An empirical approach in predicting biodiesel viscosity at various temperatures: Fuel. F.. 75-82. p. 10-6-2006. p. p. 29. F.Reference list Krisnangkura. A. and Sastry. and Sosulski. 1610-1616. p. Inc.. Mathematical simulation of an oilseed press: Journal of the American Oil Chemists' Society. Ref Type: Conference Proceeding Peter Chisawillo. 2007. V. 2005. v. C.. P. B. Faculty of Innovation Sciences. Director Vyahumu Trust. nuts and kernels. Bailey's Industrial Oil & Fat Products volume 5: "Edible oil & fat products: Processing Technologies": New Jersey. Ramaswamy. Prospects and potential of fatty acid methyl esters of some non-traditional seed oils for use as biodiesel in India: Biomass and Bioenergy.... v. Ref Type: In Press Sukumaran. 77-84. v. Singh. 1988. 2006.. Ref Type: Report Ruud van Eck. K. and Malarvizhi. 7. and Biemond. Romijn. and Bargale. K. Development of a small capacity double stage compression screw press for oil expression: Journal of food engineering.. J. Vadke. Singh. and Bargale. Eindhoven University of Technology. F.. and Pairintra. N. Sosulski. 83.Shila. 2005. 1-19. C. 28. M.. 2000. 107-113. Ref Type: Personal Communication Pramanik. v. S. R.. 444-451. A. p. 2-22-0007.. John Wiley & Sons. 239-248.. v. 54 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . 1989. P. 42. K. S.. v. v. C.. Forest college and research institute. 65. M... 2000. P. Compression of a bed of rapeseeds: The oilpoint: Journal of Agricultural Engineering Research. 2003. 65. Biosystems Engineering . and Shook. p. Waris. W. and Singh. Properties and use of jatropha curcas oil and diesel fuel blends in compression ignition engine: Renewable Energy. T. Director Intermech. TNAU. Openshaw. K. D. 106-110. Reeja.. S. Yimsuwan. Ref Type: Personal Communication Mohibbe Azam. Mechanics of oil expression from Canola: Journal of the American Oil Chemists' Society. p. V. Physical and mechanical properties of Jatropha curcas L. 293-302. fruits. 43. K. Thermal conductivity of selected liquid foods at elevated pressures up to 700 MPa: Journal of food engineering. A review of Jatropha curcas: an oil plant of unfulfilled promise: Biomass and Bioenergy. 1988. v. Mechanical expression of oil from linseed (linum usitatissimum L): Journal of oilseeds.. p. V.. S. J. I.

J. Ward. and Singh. Gas Assisted Mechanical Expression Of Oilseeds: Ph.. 80. 2005. 350-358. Willems. University of Twente. Energy analysis in the screw pressing of whole and dehulled flaxseed: Journal of food engineering. A. N.. Tostenson. Processing High Oil Content Seeds in Notinuous Screw Presses: Journal of the American Oil Chemists' Society.. Gas assisted mechanical expression of cocoa butter from cocoa nibs and edible oils from oilseeds: The Journal of Supercritical Fluids.D. 261-264. p. Wiesenborn. D. K. l.. 2007. p.. 55 Screw-pressing of Jatropha seeds for fuelling purposes in less developed countries . Wiesenborn. 2006. M. M... D. v. Feasibility study of Jatropha seed husk as an open core gasifier feedstock: Renewable Energy. Y. 1976. Vyas. K. l. and Kangas. v. P... 1039-1045.. 2007.. and Kangas. v. 2003. Zheng. A. J. P. Kuipers.Reference list Venter. B. N. Y. p. 512-517. p. D. ISBN: 978-90-365-2502-2.. 32. 66. Thesis. Screw pressing of Whole and Dehulled Flaxseed for Organic Oil: Journal of the American Oil Chemists' Society.. P. p. 37. the Netherlands. R. J. v. P. Tostenson.. 193-202. K. N. Willems. Zheng. and Haan. N..

Approximately 34 wt.2 56.3 45.1-47.40 105.5 21.5-1.5 9.15 ±0.6-4.2 25.3 25. Striking is the much lower viscosity encountered by Akintayo compared to Agarwal (17.6-27.3 40. Table A-7-2 Composition of Jatropha seed and press cake (Gubitz. 2000.5 press cake 56.0 0. shoe polish.1 4 Moisture content (wt% w.2 kernel 22.4 2.85±1. varnish etc.9 10.1 vs 35.1 shell 4.0-0.8-58.0-18. 2000.4 74.2 30. Recovery percentage of jatropha oil is expressed with respect to the whole fruit weight.468 0.) 15 8 5 15 10 3 5 5 0 3 Energy value (MJ/kg) 15.b.Appendix A: Properties of Jatropha seeds and oil Different parts of the Jatropha plant can be used as either solid or liquid fuel. 2004.4 18.3 3. The energy content and recovery percentage of several Jatropha products are shown in Table A-1.1 1.1 198. and Trabi.8 1.7-51.9-89. 2000. 307-310).2±0.% consists of the seed kernel (Openshaw.1 17.5-31.8 30 26.b.98 cST).5 2 Some of the physical and chemical characteristics of Jatropha oil are represented in Table A-3. The Iodine value given in indicates application in the manufacturing of alkyd resin.1 56 .4 8.6-10.% of the seed is hull and the other 66 wt.10 12 2.3-19.5±0.919 2.8 2.10 3. 1-19). Fuel wood whole fruit whole nut coat hull kernel wood charcoal hull charcoal plant oil press cake Ash content (wt% d.62 0. Table A-7-1 Energy values of varies Jatropha products and (Openshaw.10 17. 1-19) 2(Akintayo. 307-310).6-4. 2004.2-27. Mittelbach. Parameter crude protein (%) lipid (%) ash (%) neutral detergent fibre (%) acid detergent fibre (%) acid detergent lignin (%) Gross energy (MJ/kg) seed 1 24.5-3.16±0.6-78.2-11.0-1.2 29.76±0. 1-19).8±0.4 3.1 5.7 281.1 83.7 25.5 0. 7382) 1(Openshaw.5 11. 1999. Table A-3 Physico-chemical characteristics of Jatropha oil (Akintayo. The high saponification value denotes good properties for soap and shampoo production.5 19.4-63.4 48.7-7.8-6. Parameter Colour Free fatty acid (mg/g) Acid value (mg KOH/g) Saponification value (mg KOH/g) Iodine value (mg iodine/g) Mean molecular mass (g) Unsaponifiable matter (%) Refractive index (25 °C) Specific gravity (25 °C) Hydroxyl value Acetyl value Viscosity (30°C) cST or mm2/s Jatropha oil Light yellow 1.1-0.0 0.) 1 6 4 13 5 3 3 15 < 0.3-4.4 4.1-9.4-3.1 Recovery percentage 95-100 95-100 67-70 28-30 23-24 44-46 15-25 15-25 11-18 29-35 The composition of different components of the seed/nut are Table A-2 gives an indication of the composition of seed and press cake.

Appendix B: German quality standard for Pure Plant Oil 57 .

1kW Figure C-1 Visualization of screw dimensions (Eggers. Figure C-2 Kistler 6152 mold cavity pressure sensor. Broeck. and Stein. Angle φ is visible in the left image of Figure C-1. 2006.6mm Screw length 160mm (groove).6mm bottum Angle φ 8 degrees Cage diameter 50mm Cage length 195mm Other specifications RPM 0-70 Engine nominal power 1. 494-499) Pictures of the pressure sensor are visible in Figure C-2. Table C-1 BT50 specifications Srew dimensions Screw diameter 48. Channel height and width are indicated by H and W in the right side of Figure C-1. 11. 173 total length in contact with seed Channel height H (constant) 11. amplifier and mounting method (Kistler 6152 user manual) 58 .Appendix C: Data on BT 50 bio press setup Dimensions and other specifications of the BT50 biopress are shown in Table C-1.5mm Channel width W 11mm Number of parallel groves (6 full windings) 7 Flight thickness 5.75mm top.

110 Thermocouple values (degrees Celsius) 100 90 Thermocouple 1 Thermocouple 2 Thermocouple 3 Thermocouple 4 Thermometer 80 70 60 50 40 40 50 60 70 80 90 100 Thermostatic bath temperature (degrees Celsius) Figure D-2 Thermocouple calibration graph Temperature (degrees Celsius) 120 59 . Results are shown in Figure D-2. This proves that values from the variable speed drive are reliable. while for higher temperatures 2.5% was found. The power indicated is around 2200-2300 W. Accuracy at low temperatures was within 1%. Power 2500 140 2000 Temperature Power output (W) 1500 100 80 1000 60 500 40 20 0 0 140 0 20 40 60 80 100 120 Time (sec) Figure D-1 Maximum power peak at press jamming The type J thermocouples were calibrated using a thermostatic bath and a liquid mercury thermometer.Appendix D: Maximum power output and thermocouple calibration The peak in Figure D-1 shows the Altivar 31 power output signal in Watts at the moment the BT50 is jammed. which is twice the nominal power of the electric motor as predicted by the Altivar 31 producer.

Temperature profile heating 120 100 T (cintegrade) 80 60 40 20 0 0 20 40 60 t (min) 80 100 120 Figure E-1 equipment used for the determination of moisture content from left to right: oven. balance (0. 60 .01 g accuracy). Equipment used for the determination of moisture content is shown in Figure E-1. filter paper instead of extraction thimble.Appendix E: Measurement equipment in Tanzania The figures below give the reader an impression of the measurement equipment used for soxhlet extraction and the determination of moisture content during experiments in Tanzania. oven drying profile. soxhlet setup. Equipment used for soxhlet extraction is shown in Figure E-2. Figure E-3 shows that the accuracy of the determination of foot levels was mainly due to the type of storage container available. Figure E-2 equipment used for soxhlet extraction from left to right: mortar and pestle to grind samples before extraction.

Figure E-3 Left: oil containers. Right: determination of oil volumes and foot levels performed by Diligent co-workers. 61 .

MarelliMotori. Right top: worm positioned in the bottom part of the cage. 62 . This workshop produces several machines for agricultural applications. Rotational speed of the Sayari expeller was 55RPM powered by a standard 5. Production of the press took place at the Intermech workshop (Morogoro. The wheel on the left is for manual adjustment of the outlet restriction. One can clearly see that transport sections and pressure cones are alternately positioned along the worm shaft. Right bottom: separate worm with clear distinction between cones and transport screw.Appendix F: Sayari expeller and Intermech workshop Designs of the Sayari screw and press cage are shown more clearly in Figure F-1. Figure F-2 Assembly of Sayari expeller at Intermech workshop and final product. Italy). Tanzania). The assembly and final product look as shown in Figure G-2. Vicenza. Left: the press cage made from separate lining bars.5kW electric motor (MA132SA4. Figure F-1 Detailed views of the Sayari expeller.

3% In addition Crushed Preheating in water 80° C 63 . Limitations were encountered by trial and error and were determined by the operational restrictions of both presses.6% 7.30% In addition Crushed Torrefaction (100.50% 9. 200 ° C) Preheating in water (70 & 80 ° C) Moisture content 6.14% 5. BT50 (Netherlands) Speed 28RPM 42RPM 49RPM 56RPM 70RPM Restriction 7mm 9mm 12mm Sayari (Tanzania) Speed 56RPM Restriction 4. 140. A wider range of moisture contents was tested at Eindhoven University as better equipment was available to condition the seeds.3mm 0.00% 13. Less hull fractions were tested in Tanzania because sample sizes were to big (1020 kg) for manual de-hulling.7% 9.6mm Hull fraction 60% 100% Hull fraction 66% 80% 90% 100% Moisture content 2.0% 8.Appendix G: Overview of variables included in this research The table below shows all the variations included in the present study. The temperature ranges for pre-heating the seeds were restricted to the point where no visible alteration of the structure of the kernel was observed.5mm 3.70% 7. The moisture content of these pre-heated samples was not determined.7mm 1.30% 8.40% 6.7mm 2. 120.6% 8. 160.30% 6. 180. The table shows that the Sayari expeller could only be operated at a single speed.

64 .Appendix H: Calculation of oil recovery from soxhlet extraction The following equations are required to calculate how much oil is recovered from the seeds after pressing. (Oo+S) the sample weight before soxhlet extraction. The S can be crossed out as the amount of oil is compared per gram of solid material. Y = 1− O0 + S Os + S O0 amount of oil in Jatropha seed sample (g) amount of oil in Jatropha press cake sample (g) amount of solid material present in the sample oil content in original (g/g) oil content in sample (g/g) oil recovery (or oil yield) (g/g) Os and Oo are determined through soxhlet extraction. W0 = with: Oo Os S Wo Ws Y O0 Os Os . The starting formulas can be rewritten into the following expression:   Ws     1 − Ws  ⋅ S    Y = 1−    Wo     1 − Wo  ⋅ S     The S can be removed from the formula because the amount of solid material before and after pressing is assumed equal. Ws = . The difficulty is that Os en Oo were determined from samples with differing weight and can therefore not directly be inserted to the formula for Y. The residual oil content in the press cake is compared to the oil content in the seeds before pressing.

power and pressure.0% moisture 13.0% moisture 13. The temperature graph in Figure I-1 shows only slight fluctuations in the range of ±5%.5% moisture 9. 70RPM.3% moisture 7. 400 300 Power (Watt) normal (6. 100% hull) 65 . Average values of graphs similar to the ones in this appendix were used to construct the results in chapter 5. 100% hull) The power data closely resembles pressure data as can be seen when comparing Figure I-2 and Figure I-3.2% moisture 8.3% moisture 7. 70RPM.Appendix I: Initial measurement data The graphs represented below give an indication of the accuracy of the output signals for temperature.3% moisture Time (min) Figure I-1 Temperature profiles for different moisture contents at standard conditions for BT50 press (9mm.3% moisture 200 100 0 -2 0 2 4 6 8 10 12 14 16 Time (min) Figure I-2 Power requirement at different moisture contents at standard conditions for BT50 press (9mm. The output signals of power and pressure are more turbulent and can vary about ±10%.7% moisture) 6. 95 90 85 Pressure (bar) 80 75 70 65 60 55 50 45 -2 0 2 4 6 8 10 12 14 normal (6.5% moisture 9.2% moisture 8.7% moisture) 6.

70RPM.3% moisture 7.140 130 120 110 Pressure (bar) 100 90 80 70 60 50 40 30 20 -2 0 2 4 6 8 10 12 14 normal (6.3% moisture Time (min) Figure I-3 Smoothed pressure curves for different moisture contents at standard conditions for BT50 press (9mm.2% moisture 8. 100% hull) 66 .7% moisture) 6.0% moisture 13.5% moisture 9.

‘c’ and ‘o’ refer to seeds. Subscripts ‘s’. ash. To − M s C p . s Ts ) / M s Because of the low temperatures.oT0 + Qconv + Qcond + Qrad where M is the mass flow in kg/s. Table J-1 Composition of Jatropha seed and press cake used to calculate Cp Component Crude fat Crude protein Crude fibre Moisture Ash Carbohydrate (by difference) Jatropha seed 0.08 cake 0. 2005.c Tc + M o C p . which is commonly used for extrusion processes. Balasubramaniam.549m p 1. ‘p’.06 0.10 0. radiation heat losses are negligible compared to convective and conductive heat losses and are therefore not taken into account. s Ts + W = M c C p . One method to approximate Cp-values is by using the following expression (Zheng and others. fat. 2005.424mc + 1. SME denotes the amount of energy that remains within the products and is calculated by rearranging the energy equation (Zheng and others. T temperature in °C and W the mechanical energy input to the screw in J/s.14 The amount of energy needed per amount of seeds is expressed by the term ‘specific mechanical energy (SME). The Q terms indicate heat losses through convection.01 0. 2007.05 0.837 ma + 4. and ‘t’ water. 193-202): SME = (W − Qconv − Qcond − Qrad ) / M s = ( M c C p . The calculation of specific heats of seeds and cake is based on the composition shown in Table J-1.Appendix J: Calculation of energy requirement The energy equation for the BT50 press provides insight in the amount of energy that is needed to separate the oil form the solids. and Sastry. T and W are easily measured.675m f + 0. The value for Jatropha oil was based on values for Canola (Ramaswamy.187 mt where m is the mass fraction on a wet basis and the subscripts indicate the following: ‘c’. This value is compared to the value measured by the volt output of the Altivar 31 (variable speed drive) that regulates the power supplied to the BT50’s electric motor. carbohydrate.6 0. conduction and radiation in J/s.25 0. 67 . ‘f’.c Tc + M o C p .o . For proper analysis of the energy balance of pressing Jatropha seeds in the BT50 screw press the specific heat values have to be properly estimated. The steady state energy balance is as follows: M s C p . Comparing the required energy input to the energy content of the oil results in an indication of the energy efficiency of the mechanical extraction process. protein. M. ‘a’. The calculations should be interpreted as a rough calculation due to the use of natural materials like the Jatropha seeds. 444-451). press cake and oil respectively.085 0.47 0.065 0. Cp is the specific heat value in J/kgK. 193-202): C p = 1. Properties like oil yield and specific heat value can easily vary per batch.1 0. In addition the mass balance equation is: Ms = M0 + Mc Solving the energy equation results in an approximation for W.

193-202): Qd = −(k b Ab + k s As ) dT dl with: Qd = conductive heat loss kb = ks = thermal conductivity of screw and barrel which are assumed to be made of stainless steel Ab and As = barrel and screw outer surface area dT/dl = heat loss per unit of length Convective heat losses are approximated by the following equation also taking into account the variation in outside diameter along the press (Zheng and others. For comparisons sake calculations using known Cp-values of cottonseed result in 97kJ/kg as can be seen in Table J-2. No effort was therefore made to calculate power requirements for the Sayari expeller.25 with: h = convective heat loss Ths = hot surface temperature Tr = reference temperature /room temperature D = Outer diameter of hot surface The actual values used for estimating required energy input for screw pressing are shown in Table J-2. This might be due to inaccurate Cp-values or limitations of the model.32 hs   D  0. 2005. 2005. The value measured by the Altivar 31 is 112kJ/kg for normal seeds and 303kJ/kg for seeds with 2. 193202):  T − Tr  h = 1.14% moisture.The conductive heat losses can be estimated using Fourier’s law for steady state onedimensional heat conduction (Zheng and others. The model of Zheng et al does not approach the values measured during testing. 68 . The calculation done using previously estimated Jatropha values show a power requirement of 170 kJ/kg seed for normal seeds.

25 7.67 0.08 11 582 170 97 121.065 0.03 0.33 1 1630 1910 1556 1494 1787 2176 394.14% moidture 0.0 168337 94882 0.9 369.5 298.0019 125 23 1173 16.33 1 1630 1910 1556 1494 1787 2176 371.9 25 7.0019 343 63 3218 98 85 25 7.0 200806 124771 16.67 0.8 358.3 0.79 7.065 0.9 96.61 0.Table J-2 Input values and results for the calculation of power requirement for Jatropha pressing in the BT 50 press Mcake (kg/s) Moil (kg/s) Mseed (kg/s) Cpjatrophacake (J/kgK) Cpjatrophaoil (J/kgK) Cpjatrophaseed (J/kgK) Cpcottoncake (J/kgK) Cpcottonoil (J/kgK) Cpcottonseed (J/kgK) Tcake (K) Toil (K) Tseed (K) SME for Jatropha estimated Cp values (J/kg seed) SME for cottonseed Cp values (J/kg seed) Conductive heat losses k (w/mC) A barrel (m2) A screw (m2) DT/dl Conductive losses (W) Conductive losses (J/kg seed) Convective heat losses Thotsurface (degrees Celsius) Tbarrel (degrees Celsius) Treference (degrees Celsius) hpresshead hbarrel Lbarrel Dbarrel Lpresshead Dpresshead Convective heat loss (W) Convective heat loss (J/kg seed) Power requirement for Jatropha estimated Cp (kJ/kg seed) Power requirement for cottonseed Cp (kJ/kg seed) normal seeds dry seeds 2.08 16 806 205 129 69 .28 0.9 298.065 0.0093 0.065 0.3 0.03 0.0093 0.

0% 48.17 2.01 0.1 0.79 1.1% 55.6% light light normal normal light light light light light normal light light light light normal normal cracked seeds cracked cake total cracked dehulled (60%) seeds dehulled (60%) cake total dehulled light light 70 .06 0.23 0.48 1.42 0.82 2. Foot contents for oil from the BT50 seem to be somewhat lower than for the Sayari judging from both tables.87 1 1.Appendix K: Foot levels in oil from the Sayari expeller and BT50 press Foot levels of the oil produced by the Sayari expeller are shown in Table K-1. The composition of this ‘slimy’ component was not studies into further detail as filtering seemed to remove most of it from the oil.0% 35.6% 29.13 1.8% 45.94 0.32 0.07 0. For the BT50 values are provided in Table K-2.6% moisture seeds total 6.53 0.14 2.6% moisture press cake 6.57 0.9% 50.74 1.77 1.3% moisture press cake total 9.16 0.2% 32.34 0.3% 53.7% moisture seeds total 6.33 2 0.44 0.58 0.31 1.91 1.03 3.1 1. Lochem.87 1.76 0.3% 43.53 0.6% 37.7% moisture 8% moisture cake 8% moisture seeds total 8% moisture 8.7 1.3% 40.73 2.1 1 2.9% 31.79 1.2% light 39.9% normal 41.33 1. Some of the BT50 samples showed an additional ‘slimy’ component located in between oil and sediments.1% 17.52 0.2% 54.8% moisture seeds 8.13 1.91 1.52 0.68 0. Table K-1 Foot levels/sedimentation for oil from the Sayari expeller Sample description 6.6 3.7 1.61 0.77 3.725 2.75 1.4 0.7 3.4 0.85 1 1.69 0.7% 23.43 1.7% moisture press cake 6.5 3.875 1.65 2.47 1.47 2.8% moisture 9.2% 38.8 foot content colour 37.42 0.09 0.91 total (l) 1.3% 53.1% 52.95 0. Foot content in both tables proved too high for normal filtering techniques applied by Liquid Filtration Consultants (LFC Lochum BV.73 1.0% 43.68 1.63 1.85 0.5 1.0% 34.57 1. The Netherlands) and suggested by Ferchau (2000).9% 54.68 0.94 1.3% moisture seeds 9.6% moistre 6.55 0.3% moisture moist 2 seeds boiled seeds normal normal normal normal 0 turns 1 turn 2 turn 3 turn oil (l) 0.49 0.89 foot (l) 0.8% moisture press cake total 8.4 1.37 0.7% 43.3% 42.88 0.5% 36.

3% 33.3 8.6 4 2.7% 71 .5 4.5 6.7 3.6 4.5% 8.8 6.7 10.2% 37.3 4.8 3.6% 40.5% 34.3% 9.1 6.3% moisture 5.4% moisture 100° preheated C 120° preheated C 140° preheated C 160° preheated C 180° preheated C total (cm) oil (cm) foots (cm) slime' (cm) 9.5 4.7% 36.6 6 3.7 2.7 4.4% 11.4 9.5% 33.3 7.4 10.6 3.9% 12.8% 43.1 5.0% 46.9 7.7 2.8 9.8% 36.5% moisture 7.7 3.4 1.2 5.8 3.1% 33.3% 51.9 5.7 3.8 2.4 4.4 4.2 10.0% 59.7% 38.6 4.7 6.8% 43.5 2.2 7.1% 32.0% 30.Table K-2 Foot levels for oil from BT50 Sample description normal 28 RPM normal 42 RPM normal 49 RPM normal 56 RPM normal 70 RPM 100% hull 90% hull 80% hull 66% hull 13.4 2.3% moisture 9% moisture 8.6% 3.4 4.4 2.1 5.8 6.7 6.7 3 4.5 foot content slime' content total impurities 28.5 0 2.3 2.1% 37.2 6.5 6.7 2.2 6.4% 43.3% 45.3% 33.5 2.1 9.0% 37.

shown in Figure L-1.Appendix L: Reinartz & Keller presses and industrial oil recovery percentages Machinenfabrik Reinartz Neuss / Germany Phone: +49 2131 9761-0 Fax: +49 2131 9761-12 The AP03. Details on the press used for these tests are shown in Figure L-2. Two separate tests were carried out with Jatropha seeds from Tanzania. linseed. The Netherlands). Alkmaar..V. sunflower and jojoba. When the Jatropha seeds were de-hulled before pressing the amount of sediments appeared to increase. Tests were done under standard settings. Specific values on the amount of sediment have to be determined by amafiilter (amafiltergroup Alkmaar B. Figure L-1 Technical details of the AP08 expeller KEK Keller Remscheid / Germany Phone: +49-2191-8410-24 Fax: +49-2191-8628 Tests with Jatropha seeds were executed to find out whether high quality expellers would yield more oil than their Tanzanian counterparts. 72 . is particularly suited for soft oilseeds like rapeseed. The first test was under standard press settings for rapeseed and the second was optimized for Jatropha by inserting two larger pressure cones and changing the feed section for more transport capacity.

adjuster for outlet restriction Table L-1 gives an impression on the residual oil content of industrial oil expellers for various oilseeds. Figure L-3 Detailed view of some of the KEK P0101 components. The residual oil contents are comparable to those achieved for Jatropha in BT50. 73 . worm with cones for pressure build-up. From left to right: press cage made from separate bars. AP08 and P0101 presses.Figure L-2 Technical details of KEK P0101 expeller Some detailed pictures of some of the Keller press components are shown in Figure L-3 to underline the quality difference with the Sayari expeller shown in Appendix F. Sayar. This finding indicates that Jatropha processing should be profitable.

Table L-1 Capacities and residual oil contents for three press types applied in industrial oil mills 74 .

Appendix M: CBA results for different process setup
Decentralized processing case 1: 5 x KEK P0101 (70kg/hr) 8 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$537,977 -$537,977 1.000 -$537,977 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.925 $22,655 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.856 $20,962 $19,395 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.792 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $24,486 0.733 $17,945 $364,076 -$62,271 -$4,071 -$47,205 -$6,348 $144,382 $388,561 0.678 $263,471 NPV -$193,550 IRR -2%

case 2: 5 x KEK P0101 (70kg/hr) 24 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$567,977 -$567,977 1.000 -$567,977 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.925 $197,227 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.856 $182,482 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.792 $168,840 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $213,163 0.733 $156,218 $379,076 -$261,539 -$13,436 -$105,569 -$12,696 $606,403 $592,238 0.678 $401,578 NPV $538,367 IRR 34%

case 3: 5 x KEK P0101 (70kg/hr) 8 hours operation per day with sedimentation (oil recovery is 0.2 instead of 0.3) Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$282,003 -$282,003 1.000 -$282,003 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.925 -$13,442 -$12,437 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.856 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.792 -$11,507 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 -$14,528 0.733 -$10,647 $189,548 -$62,271 -$4,071 -$47,205 -$6,348 $105,368 $175,020 0.678 $118,675 NPV -$211,360 IRR -14%

case 4: 5 x KEK P0101 (70kg/hr) 24 hours operation per day with sedimentation (oil recovery is 0.2 instead of 0.3) Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$312,003 -$312,003 1.000 -$312,003 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.925 $46,749 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.856 $43,254 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.792 $40,020 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $50,526 0.733 $37,028 $204,548 -$261,539 -$12,214 -$12,696 -$105,569 $442,545 $255,073 0.678 $172,957 NPV $28,005 IRR 11%

case 5: 5 x Sayari (70kg/hr) 8 hours operation per day with cyclone filter and cheaper storage Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$299,575 -$299,575 1.000 -$299,575 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.925 $27,789 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.856 $25,711 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.792 $23,789 -$62,271 -$4,071 -$47,205 -$800 $144,382 $30,034 0.733 $22,011 $182,029 -$62,271 -$4,071 -$47,205 -$800 $144,382 $212,063 0.678 $143,793 NPV -$56,481 IRR 3%

case 6: 5 x Sayari (70kg/hr) 24 hours operation per day with cyclone filter and cheaper storage Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$329,575 -$329,575 1.000 -$329,575 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.925 $198,357 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.856 $183,528 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.792 $169,807 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $214,384 0.733 $157,113 $197,029 -$261,539 -$12,214 -$105,569 -$12,696 $606,403 $411,413 0.678 $278,966 NPV $658,196 IRR 63%

75

case 7: 5 x Sayari (70kg/hr) 8 hours operation per day with cloth filter and cheaper storage Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$51,850 -$51,850 1.000 -$51,850 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.925 -$8,308 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.856 -$7,687 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.792 -$7,113 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.733 -$6,581 -$62,271 -$4,071 -$47,205 -$800 $105,368 -$8,980 0.678 -$6,089 NPV -$87,628 IRR -

case 8: 5 x Sayari (70kg/hr) 24 hours operation per day with cloth filter and cheaper storage Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$81,800 -$81,800 1.000 -$81,800 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.925 $46,749 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.856 $43,254 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.792 $40,020 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.733 $37,028 -$261,539 -$12,214 -$105,569 -$12,696 $442,545 $50,526 0.678 $34,260 NPV $119,510 IRR 55% Centralized processing 0 1 2 3 4 5
case 9: 1 x Reinartz 14/30 (350kg/hr) 8 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$250,760 -$250,760 1.000 -$250,760 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.925 $15,945 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.856 $14,753 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.792 $13,650 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $17,233 0.733 $12,630 $165,518 -$62,271 -$10,178 -$53,299 -$3,624 $146,606 $182,751 0.678 $123,918 NPV -$69,865 IRR 0%

Year

case 10: 1 x Reinartz AP 14/30 (350kg/hr) 24 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$310,760 -$310,760 1.000 -$310,760 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.925 $181,915 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.856 $168,315 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.792 $155,732 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $196,614 0.733 $144,090 $195,518 -$261,539 -$30,535 -$119,808 -$7,247 $615,744 $392,132 0.678 $265,892 NPV $605,185 IRR 61%

case 11: 1 x Reinartz AP15/45 (750kg/hr) 8 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance Revenues Total Discount factor Dicounted Cash flow -$361,686 -$361,686 1.000 -$361,686 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.925 $84,491 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.856 $78,175 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.792 $72,330 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $91,318 0.733 $66,923 $211,240 -$124,543 -$20,357 -$53,299 -$3,695 $293,211 $302,558 0.678 $205,155 NPV $145,387 IRR 20%

case 12: 1 x Reinartz AP15/45 (750kg/hr) 24 hours operation per day with cyclone filter Year 0 1 2 3 4 5 Equipment Seeds Fuel Labour Maintenance -$481,686 -$523,079 -$61,071 -$153,877 -$10,124 -$523,079 -$61,071 -$153,877 -$10,124 -$523,079 -$61,071 -$153,877 -$10,124 -$523,079 -$61,071 -$153,877 -$10,124 $271,240 -$523,079 -$61,071 -$153,877 -$10,124 Revenues $1,231,487 $1,231,487 $1,231,487 $1,231,487 $1,231,487 Total Discount factor Dicounted Cash flow -$481,686 1.000 -$481,686 $483,337 0.925 $447,203 $483,337 0.856 $413,770 $483,337 0.792 $382,837 $483,337 0.733 $354,216 $754,577 0.678 $511,654 NPV $1,627,994 IRR 99%

76

Appendix N: Calculation of the Engaruka case
Poor yield 100 acres Year Equipment Running costs Revenues 0 -$4,795 1 -$7,314 2 -$7,314 3 -$7,314 4 -$7,314 Total $13,095 $13,095 $13,095 $13,095 -$4,795 $5,781 $5,781 $5,781 $5,781 Discount factor Discounted cash flow 1.000 -$4,795.06 0.925 $5,348.71 0.856 $4,948.73 0.792 $4,578.67 0.733 $4,236.28 NPV $14,317.33 IRR 115%

Standard situation 100 acres Year Equipment Running costs Revenues 0 1 -$2,450 2 -$2,450 3 -$2,450 4 -$2,450

Total $6,300 $6,300 $6,300 $6,300 $0 $3,850 $3,850 $3,850 $3,850

Discount factor Discounted cash flow 1.000 $0.00 0.925 $3,562.10 0.856 $3,295.73 0.792 $3,049.28 0.733 $2,821.25 NPV $12,728.36

Poor yields 50 acres Year Equipment Running costs Revenues 0 -$4,795 1 -$4,002 2 -$4,002 3 -$4,002 4 -$4,002

Total $6,344 $6,344 $6,344 $6,344 -$4,795 $2,343 $2,343 $2,343 $2,343

Discount factor Discounted cash flow 1.000 -$4,795.06 0.925 $2,167.51 0.856 $2,005.42 0.792 $1,855.46 0.733 $1,716.71 NPV $2,950.03 IRR 33%

Optimistic yields 100 acres Year Equipment Running costs Revenues 0 -$4,795 1 -$23,132 2 -$23,132 3 -$23,132 4 -$23,132

Total $40,097 $40,097 $40,097 $40,097 -$4,795 $16,965 $16,965 $16,965 $16,965

Discount factor Discounted cash flow 1.000 -$4,795.06 0.925 $15,696.24 0.856 $14,522.49 0.792 $13,436.51 0.733 $12,431.73 NPV $51,291.91 IRR 353%

Standard situation 100 acres optimistic yield Year Equipment Running costs Revenues 0 1 -$2,450 2 -$2,450 3 -$2,450 4 -$2,450

Total $18,900 $18,900 $18,900 $18,900 $0 $16,450 $16,450 $16,450 $16,450

Discount factor Discounted cash flow 1.000 $0.00 0.925 $15,219.88 0.856 $14,081.75 0.792 $13,028.73 0.733 $12,054.45 NPV $54,384.80

77

Appendix O: Villages for case study
Case studies Rural areas could potentially benefit from the products coming from the Jatropha plant. However most of these products, except the seeds themselves, require processing equipment for production. The example that leaps to the eye is oil production. In any case a machine to press the oil from the seeds in necessary. Storage tanks and filters are also required. Before the oil is ready to put into an engine neutralizing and filtering should be conducted, although in case of older Lister engines crude filtering is sufficient. To a lesser extent these investments also hold for soap, fertilizer and briquette production. One can imagine that most rural communities in Tanzania are unable to invest in such machinery and therefore end up selling yet another primary product at unrewarding prices. In an attempt to depict to what extend local people can exploit the benefits associated with Jatropha without a strong dependence on European and North American investors three different Jatropha cases are developed. The cases represent three villages in Northern Tanzania with a different level of potential to use Jatropha. Their classification is as follows: Temi ya Simba This small village of approximately 600 inhabitants close to Arusha has no electricity grid connection and for cooking the inhabitants use mostly wood, which they collect outside the village. Charcoal stoves or the more sophisticated plant oil stoves are not yet introduced to this region. Two farmers are growing Jatropha and around 100 people grow trees in hedges throughout the village and around agricultural land. During the first harvest 170 kg of Jatropha seeds were collected for Diligent and 600kg was used by the women group for soap production. According to Mr Rubanze the total Jatropha production potential for this village is around 4-6 tons/year two years from now (in case of two harvest seasons per year). Local applications for Jatropha oil are substitution of conventional diesel for the five village maize mills and soap production by a local women group supported by Kakute. It is interesting to find out if the people are interested in electricity, charcoal or even soap production on a larger scale. Error! Reference source not found. shows diagram clarifying the Jatropha applications and the flow of goods between actors in Temi ya Simba. The colours classify the benefits of Jatropha to a certain actor. Farmers are the only village members assumed to generate additional income form Jatropha. The user experience more practical advantages like increases fuel availability and healthier cooking methods due to decreased smoke production.

Figure 0-1 Jatropha utilization in Temi ya Simba

Maji ya Chai In the region of Maji ya Chai many people are enthusiastic about the introduction of oil fuelled Bosch-Siemens Protos plant oil stoves by the German NGO GTZ. One reason for the introduction of the stoves is to get people to use renewable oils instead of fossil fuels or wood. The use of cooking wood is one of the most important causes fro deforestation in Tanzania. However villagers cannot afford to pay for the oil to fuel the stove as Jatropha oil and sunflower oil are more expensive than both wood and kerosene. This village serves as a case to check if growing Jatropha solely for cooking purposes is

78

Figure 0-2 Jatropha utilization in Maji ya Chai Engaruka The village of Engaruka consists of two separate parts: northern and southern Engaruka. These savings or shares form the basis for providing credit to members. As Jatropha was already grown in Engaruka for fencing most seeds still stem from hedges. which involves pooling of voluntary savings from members in the form of shares. the farmers gain additional income by processing the seeds into oil at farm level. The importance of this market segment is underlined by a SACCOs market of nearly US$1 billion in Kenya. (Enterprising Solutions Global Consulting. An impression of the MFP is given in Figure 7-1. As can be seen in the model in Error! Reference source not found. 3 houses and a campsite are provided with energy from this platform. As can be seen in Figure 6-2 the model for Engaruka shows most economical activities as a result of Jatropha processing. Together with the diesel driven generator these machines constitute the potential Jatropha consumption in Engaruka. Over 30 farmers are growing Jatropha on small plots between 0. The MFP operates five hours a day generating roughly 6 kWh. A press cooperation consisting of maybe 5-10 people could start production of Jatropha oil and grain millers could also benefit from this cheaper locally produced fuel to power their engines. The first harvest in 2007 produced approximately 2400 kg Jatropha seeds. The investment required to setup a press-cooperation probably exceed the amount of money available to the local investors. The SACCO system is a mutual membership organization. In total 20 shops. Therefore a Multi Functional Platform (MFP) was installed by a Tanzanian NGO named TaTedo in November 2006 to combine seed pressing with electricity generation and grain milling. The total land area available for Jatropha is around 100 acres. Therefore Savings and Credit Co-operatives (SACCOs) are included in this system.feasible. Most of the 9000 inhabitants of this Maasai village make a living as farmers or keeping feedstock. The available credit is normally around three times the level of savings or shares.5-3 acres in size(2007). The most important expansions are the inclusion of electricity production and a larger group of people generating additional income from Jatropha. Most of the villagers cannot get any loans from the bank as farming is excluded from the definition of enterprises the bank uses when judging loan requests. As is the case for Temi ya Simba there is no grid connection present in the village. In addition to the grain mill connected to the MFP seven more diesel powered grain mills are present in the village. 2003) 79 .

Between the engine and the maize mill normally a generator is fitted like the blue one shown in the right hand picture that was taken in Engaruka. Lister engine and a maize mill.Figure 7-1 Left hand picture: A TaTEDO test MFP near Dar es Salaam with from left to right a Sayari screw press. Figure 7-2 Jatropha utilization in Engaruka 80 .

Sign up to vote on this title
UsefulNot useful