Fuel 195 (2017) 217225

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Process optimization and kinetic modeling of biodiesel production using

non-edible Madhuca indica oil
Chandrasekaran Muthukumaran a,, Ramachandran Praniesh a, Periyasamy Navamani a, Raghavan Swathi a,
Govindasamy Sharmila a, Narasimhan Manoj Kumar b
a
Department of Industrial Biotechnology, Government College of Technology, Coimbatore 641 013, India
b
Department of Genetic Engineering, School of Bioengineering, SRM University, Kattankulathur 603 203, India

h i g h l i g h t s

Non-edible mahua oil is utilized for biodiesel production.

RSM was used to optimize the transesterification process variables.

Catalyst, methanol and temperature had significant effect on mahua biodiesel yield.

Biodiesel production follows first-order kinetics.

Mahua biodiesel satisfied the ASTM standards.

a r t i c l e i n f o a b s t r a c t

Article history: Optimization and kinetic modeling of biodiesel production from non-edible Madhuca indica oil were
Received 10 August 2016 investigated in this study. Type of catalyst, catalyst concentration, methanol amount, and reaction tem-
Received in revised form 11 January 2017 perature were optimized by the univariate method. KOH was found as a better catalyst for conversion of
Accepted 15 January 2017
mahua oil to biodiesel. Response surface methodology (RSM) was employed to determine the optimal
level of KOH (%), methanol amount (v/v), temperature (C) and time (min). Maximum biodiesel yield
of 91.76% was predicted at the optimal level of KOH as catalyst (1.5%), methanol amount (0.32% v/v), tem-
Keywords:
perature (60 C) and time (90 min). Biodiesel yield (88.71%) was obtained in the validation experiments
Mahua oil
Transesterification
and fitted 96.6% with the RSM predicted results. Kinetic studies were performed at different temperatures
Biodiesel and observed that the conversion of mahua oil to biodiesel follows the first order reaction. The kinetic
Optimization rate constants and activation energy were calculated. The physiochemical properties of mahua biodiesel
RSM were determined using standard methods and the mahua biodiesel properties are in accordance with the
Kinetics ASTM D6751 standards.
2017 Elsevier Ltd. All rights reserved.

1. Introduction
An image on fuel consumption represents the oil demand in
India. According to the Petroleum Planning and Analysis Cell, India
is Asias third-largest country consumes about 15.48 million tons
of fuel by the year 2015. On putting the countrys track on fuel
demand, it requires 300,000 barrels per day and the need surpass
the China in an incremental growth [1]. According to U.S. Energy
Information Administration (EIA), Indias demand will be more
than double to 8.2 million bbl/d by 2040 while domestic produc-
tion will remain at 1 million bbl/d. The country mainly depends
Corresponding author.
E-mail address: biopearl1981@gmail.com (C. Muthukumaran).
on imported crude oil to meet the petroleum demand. Due to these
reasons, the need for an alternative to traditional petroleum arises.
In search of an alternative source of energy which reduces the car-
bon emission to environment and dependency on fossil fuel is the
major key factor which drives towards the production of biodiesel.
Biodiesel is mono-alkyl esters of fatty acid derived from transester-
ification reaction between oil and alcohol in the presence of cata-
lyst [2]. Biodiesel has similar properties with conventional
petroleum fuel and has several advantages over it such as low
CO2 emission, safe and easy to handle, nontoxic, high fuel effi-
ciency [3,4]. Nowadays >95% of biodiesel produced from the edible
oils by transesterification with alcohol and catalyst. It will be diffi-
cult for large scale production of biodiesel because it imbalances
the global food demand and diminishes the line between the food
and fuel as both if the fields strive to gain from the same oil

http://dx.doi.org/10.1016/j.fuel.2017.01.060
0016-2361/ 2017 Elsevier Ltd. All rights reserved.
218 C. Muthukumaran et al. / Fuel 195 (2017) 217225
resources. The use of low-cost feed stock will decrease the produc-
tion cost, research on biodiesel production from non-edible oils
such as jatropha [5], rapeseed [6], karanja [7], palm [8], mahua
[9], Hodgsonia macrocarpa seed [10], Cerbera odollam [11], moringa
[12] and rubber seed [13] are increasing nowadays.
Mahua oil is a non-edible vegetable oil obtained from seeds of
Madhuca indica tree found in many parts of India. Mahua oil is an
excellent source of fuel comparable with diesel. It possesses a high
content of free fatty acids (FFA) and, therefore it makes as a good
initiator for transesterification process and can be transesterified
into methyl esters, ethyl esters and butyl esters based on treatment
with methanol, ethanol and butanol respectively. It has been
reported that the methyl esters of mahua oil is closer to the prop-
erties of diesel and also having less carbon emission [14]. Methanol
is used rather than other alcohols due to its high efficiency,
increased thermal efficiency; low stoichiometric air to fuel ratio
(6.42:1), recovery of biodiesel from final product is easy. To con-
struct an economically attainable bioprocess, the operating param-
eters must be optimized perfectly. Using an effective experimental
design, the combined interaction of processing parameters can be
optimized to produce the desired product. Response Surface
Methodology (RSM) is a promising tool for designing the experi-
ments, constructing models, analyzing the effects of factors and
scrutinizing the optimum conditions [15]. Ghadge and Raheman
employed RSM for optimizing the acid pre-treatment parameters
for biodiesel production using mahua oil [16]. In this study, process
variables for transesterification of mahua oil were optimized using
RSM and kinetic modeling of transesterification was studied.
The present study reports an approach based on the use of
mahua oil as low-cost feed stock and the influence of various oper-
ating parameters such as catalyst type, catalyst concentration,
methanol to oil molar ratio, and temperature on the yield of fatty
acid methyl esters (FAMEs). Kinetic constant, as well as activation
energy for the transesterification reaction have also been deter-
mined at optimum operating conditions. The properties of synthe-
sized biodiesel are analyzed in order to compare with ASTM
standards.
2. Materials and methods
2.1. Materials
Mahua oil was obtained from a private oil company. Methanol,
NaOH, KOH, BaCl2, Na2CO3, K2Cr2O7 were purchased from Hi-
media, India. All the chemicals used in this study were of analytical
grade.
2.2. Methods
2.2.1. Pre-treatment and acid esterification
The crude oil is very denser, and it was heated at 60 C for an
hour to remove the traces of water, and it was filtered to remove
the dust particles. The free fatty acid profile of treated oil was ana-
lyzed, and it is given in Table 1. The oil had an initial acid value of
about 36 mg KOH/g corresponding to the FFA level of 18% which is
far behind the limit for transesterification reaction (1 and 2%).
Table 1
Free fatty acid profile of treated mahua oil.
S. No. Parameters Carbon atom number
1 FFA as ricinoleic acid 18:1
2 FFA as oleic acid 18:1
3 FFA as palmitic acid 16:0
4 FFA as lauric acid 12:0
Single step acid pre-treatment was carried out to reduce the FFA
of the oil. The reaction was carried out in a 500 ml beaker using
magnetic stirrer. The methanol was initially charged in the beaker
(0.40 v/v oil) and sulphuric acid (1% w/w oil) was added by con-
stantly stirring. The acid was thoroughly mixed with methanol
and then pre-heated oil was added. The mixture was stirred at a
constant speed at 55 C for 1.5 h. The reaction mixture was then
poured in separating funnel and allowed to settle for 2 days. The
top layer had unreacted methanol and water whereas the bottom
layer had oil and fatty acid methyl esters (FAMEs) [17].
2.2.2. Determination of acid value
The acid esterified oil is titrated against 0.1 N KOH using phe-
nolphthalein indicator to estimate the FFA content. The acid trans-
esterification was repeated for two times to achieve FFA of about
0.8% which is suitable for transesterification reaction. This acid
treated oil with less FFA% was stored and used for further experi-
ments [8].
2.2.3. Transesterification
For transesterification reaction, pre-treated oil was heated at
55 C and the methanol containing dissolved catalyst was added.
The reaction was performed in temperature controlled magnetic
stirrer at a constant speed for 1.5 h and the reaction temperature
was monitored routinely using thermometer. After this step, the
mixture was poured in separating funnel and allowed to settle
for a day. Two layers were obtained, the top layer comprised of bio-
diesel and the bottom layer comprised of glycerol. The biodiesel
layer was separated and washed with an equal amount of warm
distilled water to remove the soap contents and other impurities.
The washing was continued until the clear biodiesel without soap
contents were obtained. The washed biodiesel is dried by heating
at 100 C for an hour to remove the traces of water. The yield of
biodiesel was then calculated using following formula,
Weight of biodiesel produced g
Biodiesel yield %  100 1
Weight of oil g
2.3. Univariate optimization
The initial optimization of the different operating parameters
for biodiesel production was done one-variable at a time approach.
Type of catalyst (NaOH, KOH, BaCl2, Na2CO3, K2Cr2O7), catalyst con-
centration (1, 2, 3, 4% w/w), methanol amount (0.20, 0.30, 0.40,
0.50% v/v) and temperature (30, 40, 50, 60 C) were optimized to
obtain the maximum biodiesel yield.
2.4. Statistical optimization by RSM
In this investigation, RSM was used for the optimization of the
process variable to improve the transesterification of Mahua oil
with central composite design (CCD). CCD is a successful design
for fitting a quadratic surface model in sequential experimentation.
CCD avoids a large number of unusual design points and allows a
sensible amount of information for testing the lack of fit. The
regression model was proposed and the experimental results were
analyzed using Design expert 7 software. Catalyst, methanol
Method Units Results
IS 548 (Part-1) (RA. 2006) % 15.93
% 15.08
% 13.69
% 10.69
 C. Muthukumaran et al. / Fuel 195 (2017) 217225 219

Table 2
CCD matrix for biodiesel production from mahua oil with experimental and predicted responses.

Run order Catalyst (% w/w) Methanol amount (% v/v) Temperature (C) Time (min) Biodiesel yield (%)
Experimental Predicted
1 0.5 0.3 40 60 82.9 82.4
2 1.5 0.3 40 60 80.6 80.6
3 0.5 0.4 40 60 92.7 92.3
4 1.5 0.4 40 60 90.0 89.8
5 0.5 0.3 60 60 82.1 82.3
6 1.5 0.3 60 60 82.9 82.7
7 0.5 0.4 60 60 86.5 86.1
8 1.5 0.4 60 60 85.7 85.8
9 0.5 0.3 40 90 81.4 81.1
10 1.5 0.3 40 90 78.1 78.2
11 0.5 0.4 40 90 83.3 83.1
12 1.5 0.4 40 90 80.0 79.5
13 0.5 0.3 60 90 92.4 92.2
14 1.5 0.3 60 90 91.4 91.5
15 0.5 0.4 60 90 88.3 88.0
16 1.5 0.4 60 90 86.7 86.7
17 0 0.35 50 75 91.5 92.2
18 2 0.35 50 75 89.2 89.1
19 1 0.25 50 75 74.7 74.7
20 1 0.45 50 75 79.2 79.8
21 1 0.35 30 75 83.3 84.0
22 1 0.35 70 75 91.1 91.1
23 1 0.35 50 45 85.1 85.4
24 1 0.35 50 105 84.6 85.0
25 1 0.35 50 75 84.6 84.7
26 1 0.35 50 75 84.6 84.7
27 1 0.35 50 75 84.6 84.7
28 1 0.35 50 75 84.6 84.7
29 1 0.35 50 75 84.6 84.7
30 1 0.35 50 75 85.4 84.7

amount, temperature and time were selected as four independent Since methanol [M] used in excess the reversible reaction of Eq.
variables in the transesterification process. Consequently, the (2) can be ignored. The formation of ME from TG assumed to be fol-
quadratic surface models designed with the CCD matrixes of 30 lowed first-order reaction and the mass balance equation for
experiments covering the full design of four factors shown in mahua oil [MO] can be represented as
Table 2. The coefficients were determined by multiple regression
dMO
analysis. ANOVA (Analysis of Variance) statistics was used to verify r kMO 4
dt
the significance of the data through adequate precision ratio and F
test. dMO
k dt 5
MO
2.5. Characterization of biodiesel
The conversion MO to ME can be written as
The biodiesel properties such as cloud point, pour point, flash MO MO0 1  x 6
point, viscosity, moisture content, ash content and volatile matter
were determined according to the ASTM standards and compared where [MO0] is the initial concentration of MO and x is the conver-
with the conventional diesel standards. sion. By substituting Eq. (6) in Eq. (5), we obtain the following
expressions
2.6. Kinetic modeling dMO0 1  x
k dt 7
MO0 1  x
In transesterification process, the biodiesel is produced from
reaction between triglyceride (TG) present in the vegetable oils Z Z
x
dMO0 1  x t
and the short chain alcohols in the presence of a catalyst. The k dt 8
0 MO0 1  x 0
transesterification stoichiometry equation can be represented as
follows [18] t
lnMO0 1  xj0x ktj0 9
MO 3M $ G 3ME 2
lnMO0 1  x  lnMO0  kt 10
where MO is mahua oil; M is methanol; G is glycerol and ME is
methyl esters.
MO0 1  x
The kinetic expression are adopted from Hindryawati and Man- ln kt 11
MO0 
iam [18] and rate equation of transesterification reaction can be
written as
ln1  x kt 12
dMO
r k1 MOM3  k2 GME3 3 ln1  x
dt k 13
t
220 C. Muthukumaran et al. / Fuel 195 (2017) 217225

The rate constant of the reaction can be determined from the
slope of graph plotted between ln (1  x) vs t.
3. Results and discussion
3.1. Effect of catalyst type
The effect of different catalysts such as sodium hydroxide
(NaOH), potassium hydroxide (KOH), barium chloride (BaCl2),
sodium carbonate (Na2CO3), and potassium dichromate (K2Cr2O7)
on biodiesel yield was studied. The experiments were carried out
at a catalyst concentration of 1% by weight of oil, methanol amount
0.25% v/v of oil and temperature 55 C for 1.5 h. The results
obtained were represented in Fig. 1a. Fig. 1a showed that maxi-
mum biodiesel yield of 90.66% was obtained for KOH followed by
13.33% for NaOH, 5.3% for K2Cr2O7, 5% for Na2CO3 and no yield
for BaCl2. The result is attributed to the fact that KOH has higher
catalytic activity than other catalyst for mahua oil and it has higher
solubility in methanol thus resulting in higher biodiesel yield.
NaOH leads to saponification reaction and hence biodiesel yield
is reduced. Low biodiesel yield for K2Cr2O7 is due to that it is a
powerful oxidizing agent which oxidizes methanol to form acid
and chromium ions which lead to the formation of water that
reduces the yield. Na2CO3 is insoluble in methanol but soluble in
water, the insoluble Na2CO3 settles out as an insoluble precipitate
and hence no chemical reaction. Na2CO3 have less alkalinity and
Na+ ions form an emulsion in mixture resulting in difficult separa-
tion of biodiesel from glycerol. Almost no yield obtained for BaCl2
because it has low catalytic activity. From the obtained results,
KOH was chosen as an effective catalyst for biodiesel production
from mahua oil and hence used in further experiments to study
the effect of operating parameters.
3.2. Effect of methanol amount
The amount of methanol used for transesterification process is
very important as they affect the production cost to a larger extent.
Methanol recovery from glycerol is useful to reduce the production
cost. The effect of methanol amount from 0.2 to 0.5% v/v of oil on
biodiesel yield was studied. The experiments were carried with
KOH as a catalyst, catalyst concentration of 1% by weight of oil
and temperature 55 C for 1.5 h. Fig. 1b showed that maximum
biodiesel yield of 92% was obtained for 0.50% v/v followed by
90% for 0.40% v/v, 88.57% for 0.30% v/v, 85% for 0.20% v/v. This is

Fig. 1. (a) Effect of catalyst type on the biodiesel yield (Reaction conditions: catalyst concentration = 1% w/w; methanol amount = 0.25% v/v; temperature = 55 C,
time = 1.5 h), (b) effect of methanol amount on biodiesel yield (Reaction conditions: catalyst = KOH, catalyst concentration = 1% w/w, temperature = 55 C, time = 1.5 h), (c)
effect of catalyst concentration on biodiesel yield (Reaction conditions: catalyst = KOH, methanol amount = 0.25 v/v, temperature = 55 C, time = 1.5 h) and (d) effect of
temperature on biodiesel yield (Reaction conditions: catalyst = KOH, catalyst concentration = 1% w/w, methanol amount = 0.25 v/v, time = 1.5 h).
 C. Muthukumaran et al. / Fuel 195 (2017) 217225 221

due to the fact that increase in methanol amount increases the cat-
alyst solubility and it favours the reaction towards biodiesel forma-
tion, as the time extends, the completion of the reaction was higher
and thus results in high yield [19]. In general 3 mol of methanol is
required to carry out the transesterification reaction in the forward
direction. Increasing the methanol amount decreases the glycerol
content as equilibrium is in the forward direction towards biodie-
sel formation. From the obtained results the yield increases from
0.20 to 0.50% v/v methanol but the yield increases by just 2% from
0.40 to 0.50% v/v methanol and thus 0.40 v/v was selected as opti-
mal amount for biodiesel production from mahua oil.
3.3. Effect of catalyst concentration
The catalyst concentration varied from 1 to 4% (w/w). The
experiments were done with KOH as a catalyst, 0.40 v/v of oil
and temperature of 55 C for 1.5 h. Fig. 1c showed the effect of dif-
ferent catalyst concentration on biodiesel yield. Maximum biodie-
sel yield of 70% was obtained for 1% KOH, followed by 40% for 2%
KOH, 10% for 3% KOH and almost no yield for 4% KOH. Fig. 1c
explained that the biodiesel yield was decreased with the increase
in catalyst concentration; this is mainly due to that increase in cat-
alyst concentration after optimal level may lead to increase in vis-
cosity of reaction mixture. The increase in viscosity of the reaction
affects the mixing and the mass transfer resistance will be
increased which will lead to decrease in yield. Catalyst concentra-
tion >1% may favour the backward reaction thereby shifting the
equilibrium from right to left leading to the formation of glycerol.
The increase in catalyst concentration may also reduce the yield
due to the formation of salts of fatty acid (soap salts) by saponifi-
cation process, it also prevents the clear separation of biodiesel
from glycerol. From the observed results, the optimum catalyst
concentration was selected as 1% for biodiesel production from
mahua oil [20]. Similar results were reported by Eevera and Pazha-
nichamy in biodiesel production from cotton seed oil using KOH,
the ester content increased with increase in catalyst concentration
from 0.5% to 1.5% whereas further increase in catalyst concentra-
tion decreases the ester content [21].
3.4. Effect of temperature
Temperature is an important parameter to be optimized as
increase in temperature increases the reaction rate due to increase
in kinetic energy of particles resulting in decreased reaction time.
The effect of temperature from 30 to 60 C on biodiesel yield was
studied. The experiments were carried with KOH as catalyst,
methanol amount of 0.40 v/v of oil catalyst concentration of 1%
by weight of oil for 1.5 h. From Fig. 1d, maximum biodiesel yield
of 90% was obtained for 60 C followed by 86.66% for 50 C, 70%
for 40 C, 53.33% for 30 C. Increase in temperature increases the
reaction rate due to increase in collision among the reacting mole-
cules which accelerates the chemical reaction (transesterification)
and hence increased biodiesel yield. Also solubility of methanol in
oil increases as temperature increases and thus the contact
between the reacting molecules increases and hence increased in
biodiesel yield. The increase in yield from 50 to 60 C was just
3.34% and so the optimum temperature was selected as 50 C in
terms of economy and hence used in further experiments to study
the effect of operating parameters.
3.5. Optimization of transesterification by RSM
In the present study, based on the underlying support of CCD
the simple and interactive effects of processing variables of trans-
esterification of mahua oil were evaluated. Using CCD the opti-
mization of experiments carried out by (a) performing the
statistically designed experiment based on experimental plan, (b)
analyzing the data by ANOVA, (c) studying the interaction effects
graphically using 3D-response surface plots, and (d) validating
the model. The CCD used for optimization and obtained responses
for 30 experiments is given in Table 2 for biodiesel production from
mahua oil. The results were evaluated for the four different process
variables and based on the responses the polynomial regression
modeling was done. The experimental data were subjected to mul-
tiple regression analysis and the regression coefficients were esti-
mated. By using the regression coefficients, the second order
polynomial model (in coded values) for biodiesel yield was
obtained (Eq. (14)).
Y Biodiesel 84:74  0:78A 1:27B 1:77C  0:11D
yield %
 0:17AB 0:55AC  0:26AD  1:54BC
 1:96BD 2:80CD 1:48A2  1:88B2
0:69C 2 0:11D2 14
where A, B, C, D indicates catalyst concentration, methanol amount,
temperature and time respectively.
Table 3 represents the Analysis of Variance (ANOVA) values for
the quadratic regression and significance of regression coefficients

Table 3
ANOVA statistics for biodiesel production from mahua oil.

Source Sum of squares Degrees of freedom Mean square F value p-value Significance
Model 559.44 14 39.96 185.75 <0.0001 Significant
A-Catalyst 14.77 1 14.77 68.67 <0.0001 Significant
B-Methanol amount 38.68 1 38.68 179.82 <0.0001 Significant
C-Temperature 75.44 1 75.44 350.66 <0.0001 Significant
D-Time 0.27 1 0.27 1.26 0.2794
AB 0.46 1 0.46 2.13 0.1647
AC 4.9 1 4.9 22.75 0.0002 Significant
AD 1.1 1 1.1 5.1 0.0392 Significant
BC 37.91 1 37.91 176.24 <0.0001 Significant
BD 61.74 1 61.74 286.99 <0.0001 Significant
CD 125.72 1 125.72 584.39 <0.0001 Significant
A2 60.46 1 60.46 281.04 <0.0001 Significant
B2 96.59 1 96.59 448.98 <0.0001 Significant
C2 13.19 1 13.19 61.31 <0.0001 Significant
D2 0.32 1 0.32 1.46 0.2449
Residual 3.23 15 0.22
Lack of Fit 2.73 10 0.27 2.77 0.1365 Not significant
Pure Error 0.49 5 0.099
Total 562.67 29
R2 = 0.994 Adj R2 = 0.988
222 C. Muthukumaran et al. / Fuel 195 (2017) 217225

Fig. 2. (a) Predicted and actual and (b) perturbation plots for biodiesel production using mahua oil.

Fig. 3. 3D surface plots explaining the interaction effects of (a) catalyst and methanol amount, (b) catalyst and temperature, (c) catalyst and time, (d) methanol amount and
temperature, (e) methanol amount and time and (f) temperature and time on mahua biodiesel production.

from the CCD model recruited in the optimization process to
enhance biodiesel yield. The results to the model design indicate
that the response equation is well suited for the CCD experiment.
The High F value (185.75) of the model suggested that the model
is significant for the transesterification of mahua oil. The signal
to noise ratio in the model is measured using adequate precision
and model is considered to be desirable if the adequate precision
is >4. Thus, the quadratic model for biodiesel production shows
an adequate precision of 53.69 justifies an adequate signal for
transesterification process. The model terms ensures the signifi-
cance in the process if the p values <0.05, whereas the values
>0.1 suggests that it is insignificant. In this case, A (catalyst concen-
tration), B (methanol amount), C (temperature), interaction effect
of AC (catalyst concentration with reaction temperature), AD
 C. Muthukumaran et al. / Fuel 195 (2017) 217225
Fig. 4. Kinetic studies on biodiesel production at different temperature (a) plot of ln (1  x) vs time and (b) Arrhenius plot of ln k vs 1/T.
(catalyst concentration with reaction time), BC (methanol amount
with reaction temperature), BD (methanol amount with reaction
time), CD (reaction temperature with reaction time), A2 (quadratic
effect of catalyst concentration), B2 (quadratic effect of methanol
amount), C2 (quadratic effect of reaction temperature) have signif-
icant effect on the biodiesel yield. The R2 and Adj-R2 are the coef-
ficient of determination, which express the quality of the
polynomial model. The values of R2 (0.994) and Adj-R2 (0.988)
are close to 1.0, which highly recommends that there is high corre-
lation between the observed and predicted values. Fig. 2a repre-
sents the significance of the model from the diagnostic plot of
actual against predicted values. The software suggests that the
diagnostic plot in which the data are scatter over the mid line
shows the bowling scores and cannot be predicted precisely.
Therefore, the results observed in Fig. 2a clearly depicts that the
actual values on x- axis and predicted values on y- axis are closer
to the 45 middle line passes across the data over the whole range.
Hence the values observed from the experiments are more corre-
lated with the predicted values suggested by the model. The Per-
turbation plot shows the effect of changing each factor holding
all other at constant level (Fig. 2b). A steep slope or curvature
indicates sensitivity of the response to that factor [22]. Fig. 2b
suggested that the catalyst (A), methanol amount (B) and reaction
temperature (C) shows a good curvature indicates that such factors
223
are important process factors. The reaction time (D) shows a flat-
test curve implies that the factor time is considered to be not sig-
nificant in the biodiesel yield.
3.5.1. Surface plots
To interpret the impact of independent variables used in the
process for optimization, it is important to investigate the interac-
tive effect of the two factors on the transesterification process for
biodiesel production. The three dimensional (3D) plots were made
based on the model polynomial function to analyze such interac-
tive effects (Fig. 3af).
3.5.2. Interaction effect of methanol amount and catalyst
Fig. 3a shows the 3D plot for biodiesel production as a function
of methanol amount and catalyst concentration under experimen-
tal condition defined by CCD. It bounds with the possibilities that
highly methyl ester yields (92.7%) occur at high methanol amount
(0.4) and low level of catalyst concentration (0.5). However the
ANOVA table indicates, when the methanol amount is reduced to
0.3 (low level) and high level concentration of catalyst at 1.5%,
the biodiesel yield was still maintained to a value (91.4%) nearer
to the highest yield (run 14). Thus, catalyst concentration is the
most important factor in improving biodiesel yield. But, the high
concentration of catalyst can produce emulsion and which may
224 C. Muthukumaran et al. / Fuel 195 (2017) 217225

Table 4 in temperature increase the yield of biodiesel and at higher temper-

Estimated reaction rate constant for transesterification of mahua oil at different ature the production declines steeply. Higher the reaction tempera-
temperature.
ture can enhance the biodiesel yield by improving the
Temperature (C) k (min1) R2 transesterification process. On the contrary, the overall yield
40 0.011 0.961 decreases significantly when the reaction temperature reaches
50 0.014 0.895 above 60 C. The reason offered to explain this condition is the
60 0.014 0.984 methanol starts volatilization when the reaction temperature
70 0.014 0.949
exceeded 65 C. And thus the volatilization of methanol might
reduce the yield. It can be concluded that the reaction temperature
is a significant factor in production of biodiesel [25,26] and the opti-
Table 5
Physico-chemical parameters of mahua biodiesel with ASTM standards. mum level of reaction temperature should be <65 C. Hence it is not
surprising to achieve maximum yield of 92.7% of biodiesel at 40 C of
Parameters Method Mahua ASTM D6571-02
reaction temperature and higher value of 0.4 of methanol amount.
Biodiesel
Density, kg/m3 890 860900
3.5.6. Interaction effect of time and methanol amount
Viscosity, cSt ASTM 445-72 5.9 1.96
Flash point, C ASTM D93 130 >130 As discussed earlier, higher the molar ratio of methanol more
Pour point, C ASTM D 97/57 3 the conversion of oil to methyl esters. Fig. 3e clearly showed that
Cloud point, C ASTM D 97/57 11 methanol amount at high level (0.4) and time at 60th min
Ash content, ASTM D 482 0.49 <0.02
improves the biodiesel yield to maximum of 92.7%. As the time
% (dry basis)
Moisture content, % ASTM D 3172-75 0.02 <0.05
extends to 75 min and the methanol amount above 0.4 greatly
Volatile matter, ASTM D 3172-75 50 declines the yield of biodiesel to 79.2%. Thus, it implies that exces-
% (dry basis) sive methanol increases the solubility of the by product (glycerol)
Acid value 0.30 <0.8 and which in turn increase the separation difficulty of biodiesel
[27]. As the time is insignificant to the biodiesel production, the
leads to phase separation [23]. It was observed that the interaction decrease in biodiesel yield to 79.2% is hardly due to the methanol
effect of catalyst and methanol amount is not significant. amount in sub optimal level. It was concluded that optimum level
of methanol amount is a important factor to be considered in the
3.5.3. Interaction effect of catalyst and temperature yield of methyl esters [28,29].
The effect of reaction temperature (C) and catalyst (A) on the
biodiesel yield was shown in 3D response surface plot (Fig. 3b), 3.5.7. Interaction effect of temperature and time
where the actual factors such as reaction time and methanol The 3D surface plot (Fig. 3f) showed that the interaction effects
amount were kept constant at 75 and 0.35 respectively. At low of time and temperature was studied while keeping the catalyst
level of catalyst amount (0 and 0.5), the increase of reaction tem- concentration and methanol amount to a fixed level. The maxi-
perature up to 40 C, greatly improves the biodiesel yield to a max- mum biodiesel yield of 92.7% was achieved at reaction tempera-
imum of 92.7%, but further increase in catalyst concentration ture of 40 C and 60th min. It was observed that on increasing
beyond 0.5% and temperature above 40 C, it was found stepwise the reaction temperature with time there was increase in yield of
decline in biodiesel yield. This may be due to high catalyst concen- biodiesel. Several investigators have been reported that increase
tration with high reaction temperature that favours the triglyc- in temperature influence the higher biodiesel yield [23,30]. The
erides saponification side reaction and forms soap which highly present study observed that increase in temperature lowers the
affects the biodiesel yield [24]. biodiesel yield and this may be due to evaporation of methanol
beyond the 65 C temperature and hence adversely affects the
3.5.4. Interaction effect of reaction time and catalyst transesterification process [31]. Further the exposure of transester-
The percentage of biodiesel yield on interaction effect of two ification reaction at minimum temperature and minimum time
factors such as time and catalyst was studied and depicted as 3D protects the methanol from evaporation and effectively enhances
response surface plot in Fig. 3c. Other factors such as temperature the conversion process and thus the yield of biodiesel [32].
and methanol amount was kept constant at 50 C and 0.35 respec-
tively. The results indicate that the effect of time is almost insignif- 3.6. Model validation
icant interaction with catalyst concentration. When time increases
from low (60 min) to high (90 min), the constant biodiesel yield The optimal level of selected variables was identified by the
(87.4%) was observed at 0.5% catalyst concentration. It can be con- point prediction tool in Design expert software. Maximum biodie-
cluded that the time has no effect in enhancement of biodiesel sel yield (91.76%) was predicted at optimal level of Catalyst con-
yield when time and catalyst have interaction. As contrast at centration, 1.5% (w/w), Methanol amount, 0.32% (v/v),
75th min, there was a reduction in production of biodiesel and temperature 60 C and time 90 min. In order to verify the model,
similar results observed in other reports showing an increased bio- the experiments were conducted at predicted optimal levels of
diesel yield achieved at low level and high level but not in the mid the variables and the biodiesel yield 88.71% was obtained. The
level of reaction time [23]. It was observed that the low reaction experimental value had good agreement with the predicted value
time (45 min) decreased the biodiesel content and this may be and suggested that the obtained model for biodiesel production
due to mass transfer effects. Low biodiesel content at low reaction using mahua oil was well fitted with experimental data.
time might due to the diffusion limitation exists between the oil-
methanol- catalyst mixture and thus exhibit lower reaction rate 3.7. Kinetic modeling
due to immiscibility of oil. In contrast, increase in reaction time
enhance the diffusivity and improves the yield of biodiesel content. The kinetic studies were conducted at four different tempera-
ture (40, 50, 60, 70 C) to determine the reaction rate constant. A
3.5.5. Interaction effect of temperature and methanol amount plot of ln (1  x) vs time for different temperature was shown
It is possible that the on increasing the reaction temperature in Fig. 4a and suggested that the experimental data fitted well with
leads to improve the biodiesel yield. Fig. 3d suggested that increase the pseudo-first order kinetics. The rate constant (k) was deter-
 C. Muthukumaran et al. / Fuel 195 (2017) 217225
mined for each temperature and given in Table 4. Since the rate
constant depends on the reaction temperature, the activation
energy required for the transesterification process can be deter-
mined by using Arrhenius equation.
k AeEa=RT 15
where k is the reaction rate constant; A is the Arrhenius constant or
collision factor; Ea is the activation energy (J/mol), R is the gas con-
stant (8.314 J mol1 K1) and T is the absolute temperature (K).
Taking ln on both sides to the linear transformation of Eq. (15)
as follows
 
Ea 1
ln k ln A 16
R T
The Ea and A were determined from the slope and intercept of
the graph plotted between ln k vs 1/T (Fig. 4b). In this study, the
values of Ea and A for transesterification of mahua oil were deter-
mined as 22.306 kJ/mol and 60.51 min1 and the obtained values
are comparable with the previous studies on biodiesel production
using non-edible oils.
3.8. Characterization of biodiesel
The mahua biodiesel produced by transesterification reaction
was characterized to determine its chemical and physical proper-
ties and shown in Table 5. The properties of mahua biodiesel were
found to be fitted with ASTM standards except ash content which
deviates slightly.
4. Conclusion
In this work, the process variables are optimized using conven-
tional and statistical method. Optimization of transesterification
reaction of mahua oil was achieved by four-factor CCD using
RSM in 30 experimental runs. A second-order model has been
obtained to predict the mahua biodiesel yield as a function of pro-
cess variables. This study postulates that the regression coefficient
(R2) of correlations has a value of 0.9943 indicating the good fitness
of the model. ANOVA statistics explained that the catalyst concen-
tration, methanol amount and temperature have a significant
effect on the yield of mahua biodiesel. The maximum biodiesel
yield of 88.71% was obtained with KOH as catalyst (1.5%), methanol
amount (0.32% v/v), temperature (60 C) and time (90 min). The
kinetic studies were performed and the values of Ea and A for
transesterification of mahua oil were determined as 22.306 kJ/mol
and 60.51 min1 respectively. The prepared mahua biodiesel has
met the ASTM standards. Thus the production of biodiesel from
mahua oil can be scaled up to industrial scale as the oil cost is
low. Mahua biodiesel proves to be a good alternative to petroleum
fuel as it has similar characteristics with petroleum fuel and it is
cost effective.
Funding information
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Acknowledgement
Authors are greatly acknowledged The Principal, and Head,
Department of Industrial Biotechnology, Government College of
Technology, Coimbatore for their encouragement and providing
facilities to carry out this study.
225
References
[1] Swain KC. Biofuel production in India: potential, prospectus and technology. J
Fundam Renew Energy Appl 2014;4:129.
[2] Yunuskhan TM, Atabani AE, Badruddin IA, Badarudin A, Khayoon MS,
Triwahyono S. Recent scenario and technologies to utilize non-edible oils for
biodiesel production. Renew Sustain Energy Rev 2014;37:84051.
[3] Dincer K. Lower emissions from biodiesel combustion. Energy Source A
2008;30:9638.
[4] Demirbas A, Dincer K. Sustainable green diesel: a futuristic view. Energy
Source A 2008;30:123341.
[5] Teo SH, Rashid U, Taufiq-Yap YH. Biodiesel production from crude Jatropha
curcas oil using calcium based mixed oxide catalysts. Fuel 2014;136:24452.
[6] Lin L, Li X, Cui F, Zhou H, Shen X, Dong M. Transesterification of rapeseed oil to
biodiesel on CaO/a-Fe hollow fiber catalyst: optimization by response surface
methodology. Bioenergy Res 2012;5:94957.
[7] Rathore V, Tyagi S, Newalkar B, Badoni RP. Jatropha and Karanja oil derived
DMCbiodiesel synthesis: a kinetics study. Fuel 2015;140:597608.
[8] Naik M, Meher LC, Naik SN, Das LM. Production of biodiesel from high free fatty
acid karanja (Pongamia pinnata) oil. Biomass Bioenergy 2008;32:3547.
[9] Puhan SV, Vedaraman N, Boppana VBR, Sankarnarayanan G, Jeychandran K.
Mahua oil (Madhuca indica seed oil) methyl ester as biodiesel-preparation and
emission characteristics. Biomass Bioenergy 2005;28:8793.
[10] Cao L, Zhang S. Production and characterization of biodiesel derived from
Hodgsonia macrocarpa seed oil. Appl Energy 2015;146:13540.
[11] Ang GT, Tan KT, Lee KT, Mohamed AR. Kinetic studies of sea mango (Cerbera
odollam) oil for biodiesel production via injection of superheated methanol
vapour technology. Energy Convers Manage 2015;105:121322.
[12] da Silva JPV, Serra TM, Gossmann M, Wolf CR, Meneghetti MR, Meneghetti
SMP. Moringa oleifera oil: studies of characterization and biodiesel production.
Biomass Bioenergy 2010;34:152730.
[13] Onoji SE, Iyuke SE, Igbafe AI, Nkazi DB. Rubber seed oil: a potential renewable
source of biodiesel for sustainable development in sub-Saharan Africa. Energy
Convers Manage 2016;110:12534.
[14] Puhan S, Vedaraman N, Rambrahamamand BV, Nagarajan G. Mahua (Madhuca
indica) seed oil: a source of renewable energy in India. J Sci Ind Res
2005;64:8906.
[15] Myers RH, Montgomery DC. Response surface methodology: process and
product optimization using design experiments. New York: John Wiley & Sons;
1995.
[16] Ghadge SV, Raheman H. Process optimization for biodiesel production from
mahua (Madhuca indica) oil using response surface methodology. Bioresour
Technol 2006;97:37984.
[17] Helwani Z, Othman MR, Aziz N, Kim J, Fernando WJN. Solid heterogeneous
catalysts for transesterification of triglycerides with methanol: a review. Appl
Catal A 2009;363:110.
[18] Hindryawati N, Maniam GP. Novel utilization of waste marine sponge
(Demospongiae) as a catalyst in ultrasound-assisted transesterification of
waste cooking oil. Ultrason Sonochem 2015;22:45462.
[19] Encinar JM, Pardal A, Martnez G. Transesterification of rapeseed oil in
subcritical methanol conditions. Fuel Process Technol 2012;94:406.
[20] Atadashi IM, Aroua MK, Abdul Aziz AR, Sulaiman NMN. The effects of catalysts
in biodiesel production: a review. J Ind Eng Chem 2013;19:1426.
[21] Eevera T, Pazhanichamy K. Cotton seed oil: a feasible oil source for biodiesel
production. Energy Source A 2013;35:111828.
[22] Singh S, Bajaj BK. Medium optimization for enhanced production of protease
with industrially desirable attributes from Bacillus subtilis K-1. Chem Eng
Commun 2015;202:105160.
[23] Silva GF, Camargo FL, Ferreira ALO. Application of response surface
methodology for optimization of biodiesel production by transesterification
of soybean oil with ethanol. Fuel Process Technol 2011;92:40713.
[24] Vincente G, Martinez M, Aracil J. Optimization of integrated biodiesel
production. A study of biodiesel purity and yield. Bioresour Technol
2007;98:172433.
[25] Colucci J, Borrerro EE, Alape F. Biodiesel from an alkaline tranesterification
reaction of soybean oil using ultrasonic mixing. J Am Oil Chem Soc
2005;82:52530.
[26] Hanh HD, Dong NT, Okitsu K, Nishmura R, Maeda Y. Biodiesel production
through tranesterification of triolein with various alcohols in an ultrasonic
field. Renew Energy 2009;34:7803.
[27] Azcan N, Danisman A. Alkali catalyzed transesterification of cottonseed oil by
microwave irradiation. Fuel 2007;86:263944.
[28] Hsaio MC, Lin CC, Chang YH, Chen LC. Ultrasonic mixing and closed microwave
irradiation assisted transesterification of soybean oil. Fuel 2010;89:16414.
[29] Tan KT, Lee KT. Production of FAMe by palm oil transesterification via
supercritical methanol technology. Biomass Bioenergy 2009;33:10969.
[30] Jansri S, Prateepchaik ul G. Comparison of biodiesel production from high free
fatty-acid, crude coconut oil via saponification followed by transesterification
or a two stage process. Kas J Nat Sci 2011;45:1109.
[31] Ragit SS, Mohapatra SK, Kundu K, Gill P. Optimization of neem methylester
from transesterification process and fuel characterization as a diesel
substitute. Biomass Bioenergy 2011;35:113844.
[32] Abbah EC, Nwandikom GI, Egwuonwu CC, Nwakuba NR. Effect of reaction
temperature on the yield of biodiesel from neem seed oil. Am J Energy Sci
2016;3:1620.