See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317011269

Production of Biodiesel in Pilot-Scale using Locally Available Feedstock

Materials

Conference Paper · October 2012

CITATIONS READS

2 477

3 authors, including:

Marliya Ismail Sanja Gunawardena

University of Auckland University of Moratuwa
18 PUBLICATIONS 123 CITATIONS 39 PUBLICATIONS 117 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Synthesis and Characterization of Nanosilica from Paddy Husk Ash View project

Polygodial extraction from native plant using emerging techiques View project

All content following this page was uploaded by Sanja Gunawardena on 19 May 2017.

The user has requested enhancement of the downloaded file.

Annual Transactions of IESL, pp. [25-32], 2012
© The Institution of Engineers, Sri Lanka

Production of Biodiesel in Pilot-Scale using Locally

Available Feedstock Materials
D.R.S. Hewa Walpita, F.M. Ismail and S.H.P. Gunawardena

Abstract: Biodiesel or fatty acid methyl ester is a fuel that can be produced from lipid sources. It
is popular as a totally renewable, nontoxic and biodegradable alternative for fossil based diesel due to
its numerous environment benefits associated with its use. Currently, biodiesel is produced mainly
using edible oils and the rest is covered by non-edible oils, animal fats and waste cooking oils (WCOs).
The lab-scale studies of the research identified that, locally available feedstock materials such as
Coconut, Palm, Rubber Seed Oil (RSO), Jatropha, Neem, Lard and WCO can be successfully converted
into biodiesel. The study used alkali-catalyzed transesterification process to produce biodiesel from
oils with FFA content less than 2.0%. The oils with free fatty acid content greater than 2.0% were pre-
treated using acid-esterification. The produced biodiesel were compatible with ASTM D 6751 and EN
14214 standards. A 50 litre portable reactor system was designed and fabricated with multi
disciplinary units based on lab-scale results for pilot-scale studies. The designed unit is capable of
carrying out complete biodiesel production process from pre-treatment to drying. The potential of
using the pilot-unit in biodiesel production was studied using Palm oil and WCO as feedstock
materials. The studies also successfully tested the quality of the biodiesel produced in its pure form
using a diesel three-wheeler.

Keywords: Biodiesel, Pilot-scale, Reactor

1. Introduction vehicle manufactures recommend only using

biodiesel blends up to 20% in their diesel
Various alternative fuels have been investigated vehicles.
to replace conventional petroleum fuels since
the first oil crisis of the 1970’s. The initial As a point of comparison, the calorific value of
interest was mainly as a supply security, but pure biodiesel (B100) is about 90% of
recently more attention has been focused on the conventional diesel, and hence, the engine
use of renewable fuels in order to reduce the performance is only reduced by 10%. Biodiesel,
CO2 emissions from combustion sources [1]. however, is made from renewable resources, is
Though, unmodified vegetable oils have the biodegradable and nontoxic, and has a higher
potential to be used as fuels for diesel engines, flash point than normal diesel. In addition,
prolong operation of engines were impossible biodiesel increases lubricity (even at blends as
even with fully refined oils. However, it was low as 3% or less), which prolongs engine life
found later that vegetable oils and fats after and reduces the frequency of engine part
transesterification with short chain alcohols replacement. Another significant advantage of
such as ethanol or methanol can be used in biodiesel is its low emission profile and its
diesel engines without any complications [2]. oxygen content of 10-11%. Additionally, it is
Since this transesterified oil could be directly also capable of reducing carbon and total
used in diesel engines it was known as particulate emissions as well as carbon
“biodiesel”. In addition to vegetable oils and monoxide (CO) and unburned hydrocarbons
animal fats, waste greases both yellow and
brown are also commonly used in biodiesel
preparation [3]. Biodiesel can be used as Eng. D. R. S. Hewa Walpita, B.Sc. Eng. (Moratuwa),
conventional diesel in diesel engines because its M.Sc. (Moratuwa),
Eng. (Dr.) (Mrs.) F. M. Ismail, B.Sc. Eng.
properties are very close to petroleum-based (Moratuwa),Grad. IChem.C (SL), Ph.D (UMIST,UK),
diesel. For example, biodiesel has the proper Senior Lecturer, Dept. of Chemical & Process Engineering,
viscosity, a high flash point, a high cetane University of Moratuwa
Eng. (Dr.) S. H. P. Gunawardena, B.Sc. Eng. (Moratuwa),
number, and therefore, researchers have found
Ph.D. (Birmingham), Senior Lecturer, Dept. of Chemical &
out no engine modifications are required when Process Engineering, University of Moratuwa.
using biodiesel [1,4]. However, still most of the

25
significantly. Biodiesel is regarded as an Oil or Fat
environmentally friendly biofuel since it
provides a means to recycle carbon dioxide Estimation of FFA
(CO2) and hence, biodiesel does not contribute
to global warming [3].
FFA < 2% FFA > 2%
Since the quality of biodiesel is a key factor
Acid Esterification H2SO4 +
when using it as an alternative to conventional
Methanol
diesel, several standards (i.e. ASTM D 6751 and
EN 14214) have been introduced. When these FFA > 2%
standards are met, biodiesel or its blends can be
FFA < 2%
used in most of the modern engines within the
manufactures recommended limits without
modifications while maintaining the engine’s Transesterification KOH +
durability and reliability. Methanol

Currently, biodiesel is prepared by Biodiesel Purification

transesterification of oils/fats using short chain (Layer separation, washing
alcohols in the presence of an alkaline catalyst. and drying)
Industrially, NaOH and KOH are preferred
with alcohols like methanol due to their wide
Biodiesel
availability and low cost [3]. If the free fatty
acid (FFA) level of oil is greater than 2%, the Figure 1 – Flow Diagram of the Biodiesel
FFA level of the oil or fat has to be reduced by Production Process
esterification to avoid soap formation in
transesterification. This is done by reacting FFA 2. Design and Construction of
with KOH or NaOH in the presence of an acid Pilot-Scale Reactor System
catalyst. The flow diagram of the biodiesel
production process is shown in Figure 1. 2.1 Process Design
The pilot-scale reactor system has been
The cost of biodiesel, however, is the main designed as a semi-automated, pilot-scale
hurdle in commercialization of the product. Use portable demonstration rig with a 50 litre
of low-cost raw-materials such as “waste reactor capacity. The designed system is capable
cooking oil” (WCO) and adaption of high of producing 20-35 litres of biodiesel per batch
quality glycerol recovery systems for the by- depending on the quality of feedstock. It is
products are some primary options which can capable of handling all the steps of the biodiesel
lower the cost of biodiesel [5]. production from feedstock pre-treatment to
biodiesel purification.
Sri Lanka as a petroleum fuel importing country
can reduce fuel importing cost if biodiesel is The reactor system was designed with 3
produced locally. When considering the multifunctional units; reactor unit, a mixing
feedstock materials, Sri Lanka is rich in wide unit and a storage unit. In addition to these 3
range of oils and fats including edible oils, non- major units, a condenser unit was attached to
edible oils, animal fats and restaurant waste the reactor and an air sparger was included to
oils. Hence, it is important to evaluate the the storage unit to fulfil additional process
feasibility of using such oils and fats in requirements. Non-compulsory units such as
biodiesel production. storage vessels which are not directly involved
to the process were omitted from the design.
Block diagram of the pilot-plant is given in
Figure 2.

2.2 Reactor Unit

The reactor unit was designed as a closed
toriconical bottomed cylindrical vessel unit to
avoid methanol evaporation during pre-
treatment and biodiesel reactions. The reactor
is equipped with an electrical heating system
and a condenser unit. Jet mixing was used as

26
the mixing mechanism since it facilitates to keep MV – MANUAL VALVE

SV – SOLENOID VALVE
the reactor as a completely closed system. The SV6

REACTOR
system was designed as a single side-entry UNIT SV5 MV 3

nozzle system with a pump and an external SETTTLING

UNIT
MIXING
circulation path with a downward pointing UNIT

nozzle at the releasing end. An additional

nozzle was placed at the middle of the reactor SV2

to make it possible to operate the reactor with SV3 SV1

smaller batches. Two control valves were used MV1

with the nozzles to select the appropriate nozzle MV2

Pipe material - SS 304L
according to the operating liquid level. SV4 Pipe size - 0.75’ scheduled 40
CENTRIFUGAL PUMP

KOH H2SO4 METHANOL

Manually controlled
Figure 3 – Piping Diagram of the Biodiesel
STORAGE STORAGE STORAGE
Full / semi automated Unit
Essential components

2.5 Unit Arrangement

MIXING

Neg lected
UNIT

components
The processing units of the plant were mounted
on a single structure and connected using two
WATER
TANK pipe lines with manual and automated valves
as necessary to carryout the overall process.
STORAGE
REACTOR

HIGH FFA
FEEDSTOCK TANK Piping diagram of the biodiesel unit is given in
UNIT
UNIT

Figure 3. Pilot plant included a semi automated

LOW FFA
FEEDSTOCK TANK control system to control the plant operations.
The final appearance of the pilot-scale biodiesel
reactor system is as in Figure 3.

CRUDE
METHANOL
GLYCERINE
TANK
WASHED
W ATER TANK
PURE BIODIESEL
STORAGE TANK
2.6 Mechanical Design
Mechanical design of the pilot plant was
carried out according to British Standard PD
Figure 2 – Block Diagram of the Biodiesel 5500, EN 13445 and ASME Code Section VIII. For
Plant the mechanical design, the design temperature
2.3 Settling Unit was taken as 125°C, and the pressure of the
Toriconical bottomed cylindrical vessel with 50 reactor vessel was taken as 29.49 kPa. SS 304L
litre capacity is used as a settling unit to was used as the fabrication material for vessels
support in non-reaction operations such as and 0.75’ scheduled 40 pipes was used for pipe
settling, layer separation and washing. A lines which are in contact with H2SO4, water
removable air sparger system consisted of a and water vapour to avoid corrosion. Carbon
mesh-like tube structure placed at the bottom of steel was used for the plant structure where it
the vessel provides tiny air bubbles during the does not have any contact with reactive
washing process. materials.

2.4 Mixing Unit 3. Materials and Method

A similar shaped mixing unit with 9.9 litre
3.1 Materials
capacity was designed to use as both methoxide
reactor and H2SO4-methanol mixing unit based Both lab-scale and pilot-scale studies were done
on the results obtained in lab-scale studies. A using locally purchased/collected feedstock oils
stirred tank agitation method was selected as and fats. In lab-scale studies coconut oil, rubber
the mixing mechanism pitched blade type seed oil (RSO), WCO, jatropha, neem and lard
turbine impeller and was powered by a AC were used as feedstock materials and the pilot-
motor. scale studies were done using WCO and palm
oil. Laboratory grade KOH, laboratory grade
H2SO4 and industrial grade methanol were used
as reactants in both lab-scale and pilot-scale
studies.

27
 separated bottom layers were settled further for
another 22 h. Then the AV of the further settled
layers was measured and the FFA amounts
were calculated. If this value is greater than 2%,
then acid esterification step was repeated and if
it is less than 2%, pretreated oil was used for
biodiesel production.

 Alkali-catalyzed transesterification
Lab-scale alkali-catalyzed transesterification
reactions were carried using 100 ml of pre-
treated oil and coconut oil samples. The
amounts of KOH require for the reactions were
estimated as given in [6]. Methanol-KOH
mixtures were prepared by dissolving
measured amount of KOH in 20 ml methanol
(20% of the oil). Then, oil samples were
preheated to 55°C, and the prepared methanol-
KOH mixtures were introduced. Reaction was
Figure 4 - Unit Arrangement of the Biodiesel
carried out for 90 min by maintaining the
Pilot Plant
temperature at 55°C while mixing. After
completion of the reactions, the mixtures were
3.2 Reaction Set-up
settled for 24 hrs and the top layers were
A 250 ml reagent glass bottle with a stopper and
separated. The separated top layers were then
a magnetic stirrer were used as a laboratory
water washed using tap water for several times
scale reactor for all the lab-scale studies. A
until clear water layers are obtained. The
thermometer was inserted to the bottle through
washed biodiesel was then heated at 110°C for
a stopper to measure the temperature of the
sufficient time to remove moisture. Washed and
reaction mixture. This complete system was
dried biodiesel samples were tested for
placed on a heating magnetic stirrer to provide
properties such as density, kinematic viscosity
heating and mixing. All pilot-scale studies were
and flash point.
done using the fabricated 50 L reactor system.
3.4 Biodiesel Production in Pilot-Scale
3.3 Biodiesel Production in Lab-Scale
As in the lab-scale studies, low FFA containing
Among the feedstock materials tested, only
palm oil was directly subjected to
coconut oil had a low content of FFA and hence,
transesterification in pilot scale studies.
directly subjected to transesterification. Acid
However, acid esterification was done prior to
esterification was done prior to alkali-catalyzed
alkali-catalyzed transesterification for WCO
transesterification as a pre-treatment step to
with a high FFA content.
reduce FFA content to a value below 2% for the
rest of the oils which had a high FFA content.
 Oil pre-treatment
AV of the WCO was measured and the amounts
 Oil pre-treatment
of H2SO4 and methanol to be used were
Acid value (AV) of the oil samples was
estimated based on the FFA present as 0.10 g of
measured according to ASTM D 1980-67 and
H2SO4/g of FFA and 2.5 g of methanol/g of
the FFA of oils were calculated. The amounts of
FFA. The same procedure followed in lab-scale
H2SO4 and methanol to be used were estimated
oil pre-treatment was adapted and the
based on the initial amounts of FFA present in
remaining FFA amount was estimated after
the samples [7]. In all esterification reactions
settling.
done in laboratory scale work, 0.05 g of
H2SO4/g of FFA and 2.5 g of methanol/g of
 Alkali-catalyzed transesterification
FFA were used. Oil samples were preheated
Alkali-catalyzed transesterification studies were
and H2SO4-methanol mixtures were introduced.
done in the 50 L reactor unit as pilot-scale
Then the reaction was carried out at
studies for palm oil and pre-treated WCO
atmospheric pressure for 30 min by maintaining
samples of 25 litres each. The required KOH
the temperature at 55°C while mixing. After the
and methanol amounts were estimated as in the
reaction time, samples were allowed to settle for
lab-scale studies. The oil samples were then
2 h and then the layers were separated. The

28
reacted with methanol-KOH mixtures for 1.5 h. All six feedstock materials tested have reached
The top layers were separated after 24 h of ASTM and/or EN standards of the flash point.
settling, and pure biodiesel samples were The required density level was also achieved by
obtained after washing and drying. Produced all biodiesel samples except coconut and
biodiesel samples were then tested for jatropha oils. Although coconut oil and RSO
properties such as density, kinematic viscosity derived biodiesel were within the standard
and flash point. values of kinematic viscosity, WCO, Jatropha
and Neem derived biodiesel have higher
3.5 Engine Testing and Road Testing viscosities. However, there is a high viscosity
Biodiesel was tested for its performances using reduction in all biodiesel samples when
70 kW “Toyo” AC diesel generator and road compared with the respective original oils. All
testing was carried out using an unmodified the kinematic viscosity readings were measured
“Piaggio” diesel three-wheeler. at room temperature (i.e. within 27-30°C) and a
low values can be expected if these
4. Results and Discussion measurements were done at the standard
measuring temperature of 40°C. Further,
4.1 Biodiesel Production in Lab-Scale incomplete transesterification reaction could be
The properties of biodiesel produced using six another reason for this high viscosity.
different oil/fat feedstocks are given in Table 1.

Table 1 - Properties of Biodiesel Produced in Lab-Scale using Different Feedstock Materials

Feedstock material ASTM/EN

Property Unit
Coconut RSO WCO Jatropha Neem Lard
°C 110 126 172 155 160 179 > 130/110
Point
Density g/cm3 0.835 0.877 0.877 0.955 0.881 0.868 0.86-0.9
cSt 4.9* 5.7* 10.26* 17.12* 8.51* - 1.9-6.0
Viscosity
* Measured at room temperature

4.2 Biodiesel Production in Pilot-Scale 4.3 Energy Consumption for Biodiesel

The properties of biodiesel produced using Production
WCO and palm oil are given in Table 2. The energy consumption in biodiesel
According to the results obtained, flash points production in the pilot-scale unit is reported
and densities were compatible with ASTM and here as it is an important parameter which
EN standards. However, kinematic viscosities affects the production cost when produce in
have shown higher values than the standards, commercial-scale. The total energy requirement
as observed in the lab-scale studies. When the of the unit is equal to its electrical energy
properties of biodiesel produced in pilot-scale requirement, since it is completely powered by
and lab-scale are compared, no significant electricity. The energy consumption was
difference was observed. This confirms that measured using the rated power consumptions
scaling up has been successful with minimum and the running times of process equipment.
product deviations. Reasonable rated powers were assumed for
equipment whenever rated powers were not
available. The summary of assumed/used rated
Table 2 - Properties of Biodiesel Produced powers of process equipment is given in
using WCO and Palm Oil in Pilot-Scale Table 3.
Table 3 – Rated Power Summary of Process
WCO Palm oil
Equipment
Property Unit derived derived
biodiesel biodiesel Rated Power
Process Equipment
Flash Point °C 179 192 (W)
density g/cm3 0.868 0.869 Circulation pump 370
Kinematic Control system 50
cSt 8.06 8.63
Viscosity Heaters (per unit) 2,000
Mixing motor 25
Air/Oxygen pump 30
Solenoid valve 20

29
According to the estimation, the energy The material cost for the production of biodiesel
required to produce 35 litre biodiesel batch was using WVO was estimated based on the results
8.702 kWh (Table 3). Hence, the energy required of both lab-scale and pilot-scale studies. The
to produce one kg of biodiesel is 0.287 kWh price of WVO was taken as Rs. 20.00 per litre
(1,032 kJ/kg), and is at an acceptable level when and the cost of the rest of the chemical required
compared with the calorific value of biodiesel were their bulk market price. According to the
(which is 35-40 MJ/kg). estimation, cost of material was Rs. 53.51 for
WVO with 5.595% of FFA (Table 5). However,
the price will vary with the FFA content of the
oil.
4.4 Raw Material and Chemical Cost
Table 4 - Electrical Energy Consumption for Biodiesel Production per Batch
Electrical energy
Number Running Rated
Production stage Equipment consumption
of units time (min) Power (W)
(kWh)
Oil Pre-treatment Stage
H2SO4 and Mixing motor 1 5 25 0.002
methanol mixing Control system 1 5 50 0.004 0.006
Electrical heaters 3 7 2,000 0.700
Circulation pump 1 7 370 0.043
Oil preheating
Solenoid valves 2 7 20 0.005
Control system 1 7 50 0.006 0.754
Electrical heaters 3 10 2,000 1.000
Circulation pump 1 30 370 0.185
Reaction stage
Solenoid valves 2 30 20 0.020
Control system 1 30 50 0.025 1.230
Solenoid valves 2 5 20 0.003
Layer separation
Control system 1 5 50 0.004 0.008
1.997
Biodiesel Production Stage
Methoxide Mixing motor 1 5 25 0.002
reaction Control system 1 5 50 0.004 0.006
Electrical heaters 3 7 2,000 0.700
Circulation pump 1 7 370 0.043
Oil preheating
Solenoid valves 2 7 20 0.005
Control system 1 7 50 0.006 0.754
Electrical heaters 3 30 2,000 3.000
Circulation pump 1 90 370 0.555
Reaction stage
Solenoid valves 2 90 20 0.060
Control system 1 90 50 0.075 3.690
Circulation pump 1 5 370 0.031
Layer separation Solenoid valves 2 5 20 0.003
Control system 1 5 50 0.004 0.038
Biodiesel Purification Stage
Oxygen pump 1 30 60 0.030
Biodiesel washing
Solenoid valves 2 5 20 0.003 0.033
Electrical heaters 3 20 2,000 2.000
Circulation pump 1 20 370 0.123
Biodiesel drying
Solenoid valves 2 20 20 0.013
Control system 1 20 50 0.017 2.153
2.186
Total Energy Consumption 8.702

4.5 Engine Testing and Road Testing the results of fossil diesel for the tested 2.5 - 3.5
The load profile obtained from engine testing of kW range. The results are shown in Figure 5.
WVO derived biodiesel is in accordance with The road testing done using 100% WCO and

30
 Fuel consumption (m/kW hr)
neem derived biodiesel in unmodified diesel 1200
Biodiesel Conventiona l Diesel
engine showed similar engine ignition as fossil
diesel and was able to drive the vehicle 1000

smoothly. However, it is not recommended to 800

use pure biodiesel in engines without
identifying long-run effects. 600

400

200

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
Power (kW)

Figure 5 – Load Profile for Biodiesel and Fossil Diesel

Table 5 – Material Consumption for Biodiesel Production per Litre

Unit price Cost
Production stage Material Unit Quantity
(Rs.) (Rs.)
Material usage
WVO lit 1.1080 20.00 22.16
Oil Pre-treatment Methanol lit 0.2074 125.00 25.92
H2SO4 kg 0.0055 80.00 0.44 48.52
FFA reduced WVO lit 1.0526 0.00 0.00
Biodiesel
Methanol lit 0.2105 125.00 26.32
production
KOH kg 0.0200 260.00 5.19 31.51
Purification
Water for washing lit 5 0.10 0.50 0.50
Material recovery
Methanol (50% of total) lit 0.2089 125.00 (26.12) (26.12)
Cost of material 54.01

5. Conclusions scaling up. Studies also showed that the

material cost and the energy requirement for
Lab-scale and pilot-scales studies showed the biodiesel production in pilot-scale is Rs. 53.51
potential of using locally available oils and fats (for WCO) and 0.287 kWh/kg (1,032 kJ/kg),
such as coconut, palm, RSO, neem, jatropha, respectively. Apart from biodiesel production
lard and WCO in biodiesel production. The studies, road testing proved the ability of using
properties of biodiesel produced using different biodiesel in unmodified engines even as a pure
feedstock materials were up to the level of fuel.
ASTM and/or EN standards in their flash point
and density. However, kinematic viscosity has References
shown a considerable deviation from the
standards, though the biodiesel produced using 1. Scholl, K. W. and Sorenson, S. C., "Combustion
coconut and RSO were within the limits. The of soybean oil methyl ester in a direct injection
reasons for higher kinematic viscosity are diesel engine”. SAE International, 1993.
assumed to be the incompletion of 2. Vellguth, G., "Performance of vegetable oils and
transesterification reaction and incorrect their monoesters as fuels for diesel engines", SAE
temperature at viscosity measuring. International, 1983.
3. Lotero, E., Liu, Y., Lopez, D. E., Suwannakarn,
A 50 litre pilot plant which could handle K., Bruce, D. A. and Goodwin, J. G., "Synthesis of
pretreatment, processing and washing of biodiesel via acid catalysis". Industrial &
biodiesel was designed and fabricated. The Engineering Chemistry Research 44(14): 5353-5363,
studies done in pilot-scale showed that the 2005.
quality of biodiesel produced in pilot-scale is 4. Noiroj, K., Intarapong, P., Luengnaruemitchai,
same as in the lab-scale, and no significant A. and Jai-In, S., "A comparative study of
corrections are required for the process when koh/al2o3 and koh/nay catalysts for biodiesel

31
 production via transesterification from palm oil".
Renewable Energy 34(4): 1145-1150, 2009.
5. Ma, F. and Hanna, M. A., "Biodiesel production:
A review". Bioresource Technology 70: 1-15, 1999.
6. Addison, K. (2008), "Biodiesel processors".
http://journeytoforever.org/biodiesel_processo
r.html. Accessed on 11th March 2009.
7. Walpita, D. R. S. H., Gunawardena, S. H. P.,
Ismail, F. M., Sampath, P. R. A. U. and
Samarakoon, S. P. A. G. L., Quantification of
reactants required in the conversion of free fatty
acids (FFA) present in vegetable oils and animal
fats into fatty acid methyl ester, Sri Lanka Patent
Pending, 2011.

32
View publication stats