International Journal of Green Energy

ISSN: 1543-5075 (Print) 1543-5083 (Online) Journal homepage: http://www.tandfonline.com/loi/ljge20

Esterification of free fatty acids: experiments,

kinetic modeling, simulation & optimization

Zakir Hussain & Rakesh Kumar

To cite this article: Zakir Hussain & Rakesh Kumar (2018) Esterification of free fatty acids:
experiments, kinetic modeling, simulation & optimization, International Journal of Green Energy,
15:11, 629-640, DOI: 10.1080/15435075.2018.1525736

To link to this article: https://doi.org/10.1080/15435075.2018.1525736

Published online: 10 Oct 2018.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ljge20
INTERNATIONAL JOURNAL OF GREEN ENERGY
2018, VOL. 15, NO. 11, 629–640
https://doi.org/10.1080/15435075.2018.1525736

Esterification of free fatty acids: experiments, kinetic modeling, simulation &

optimization
Zakir Hussain and Rakesh Kumar
Department of Chemical Engineering, Rajiv Gandhi Institute of Petroleum Technology, Jais, Amethi, India

ABSTRACT ARTICLE HISTORY

The development of the process which can mitigate the drawbacks of catalytic esterification and Received 4 October 2017
handles high free fatty acid (FFA) containing oils is the highly focused area in biodiesel production. In Accepted 16 September 2018
view of attaining the cleaner biodiesel production, the present research efforts are focused on studying KEYWORDS
the methyl esterification of FFA present in Karanja oil non-catalytically in a batch reactor. Kinetics of the Aspen plus; Karanja oil;
reaction was modeled as the pseudo first order in the forward direction & second order bimolecular type kinetics modeling; non-
in the backward direction to deduce kinetic parameters. The obtained parameters were used to simulate catalytic esterification;
the process in Aspen plus®. Experimental results show that 96% conversion of FFA can be achieved at simulation
220°C and 1:6 (w/v) oil to methanol ratio. The calculated activation energy and rate constant are
48.53 kJ/mol and 0.641 min−1, respectively, for the forward reaction and 18.74 kJ/mol and 4.18E−4 (g)/
(mgKOH.min) respectively, for the backward reaction. Simulation results showed a little higher conver-
sion (99.85%) of oleic acid compared to the experimentally observed conversion (96%) at similar
reaction conditions. The optimal process parameters were estimated using sensitivity analysis of
Aspen Plus along with heat integration.

1. Introduction Esterification is a reversible reaction in which equimolar

quantities of FFA and alcohol react to produce the equimolar
Biodiesel has been accepted worldwide as a green fuel and
amount of alkyl ester and water in the presence or absence of
considered as an alternative to the diesel fuel (Sara, Brar, and
the acid catalyst (Berrios et al. 2010; Kolyaei et al. 2016). The
Blais 2016; Van et al. 2016; Sebastian, Muraleedharan, and
homogeneous acid catalyzed esterification process has various
Santhiagu 2017). Considering the eco-friendly nature of bio-
disadvantages like corrosion of the equipment and difficulty
diesel such as its biodegradability, lesser greenhouse emis-
in product separation & catalyst recovery (Cho et al. 2012). To
sions, superior lubricating properties and intrinsically free of
curtail these drawbacks, heterogeneous acid catalysts such as
sulfur compared to diesel, federal and state governments are
sulfonic-functionalized organic resins, sulfonic polystyrene-
framing policies to widen its use (Axelsson et al. 2012; Torres
divinyl benzene, niobium phosphate (Bassan et al. 2013) and
et al. 2013; Long and Fang 2012).
sulfated zirconia have been used for the production of biodie-
Industrial scale biodiesel is produced using homogeneous
sel using Karanja oil containing FFA (Gupta, Kiro, and Deo
base-catalyzed transesterification of vegetable oils or animal fats
2016). The reported yield of biodiesel was 38% with fresh
due to their faster reaction rates compared to acid-catalyzed
catalyst and reduces further to 16% with reused catalyst
process (Lee et al. 2014). This process is highly sensitive towards
(Lou et al. 2012; Zong et al. 2007). The non-catalytic super-
the presence of FFA and initiates the saponification reaction
critical processes have also been explored to overcome the
which hinders the progress of the main reaction, transesterifica-
limitation of the catalytic processes and have shown signifi-
tion (Hussain, Mohammad, and Kumar 2016; Pisarello et al. 2010;
cantly higher yield (~97%) of biodiesel in a very short period
Rathore et al. 2015). Moreover, it also inhibits the separation of
of time (Gómez-Castro et al. 2011; Anitescu, Deshpande, and
products; glycerol and biodiesel, during purification and results in
Tavlarides 2008; Vieitez et al. 2009; Shin et al. 2012; Alenezi
poor quality of biodiesel (Alper Tapan, Yıldırım, and Erdem
et al. 2010). But, the severe conditions employed in the super-
Günay 2016). Therefore, it is essential to limit the FFA in feed
critical processes demand higher capital cost as well as oper-
oil prior to transesterification reaction (Stacy, Melick, and
ating cost which limits their commercialization (Du et al.
Cairncross 2014; Gandhi and Kumaran 2014). Several pretreat-
2013) and some studies (Shin et al. 2011; Quesada-Medina
ment techniques such as esterification, distillation, and neutraliza-
and Olivares-Carrillo 2011) even proved the thermal degrada-
tion are employed to lower the FFA content in feed oil. Among all
tion of esters at supercritical synthesis condition.
these techniques, esterification process is found to be the most
The non-catalytic subcritical esterification process miti-
efficient in reducing the FFA content of feed oil (Chai et al. 2014;
gated the above-stated drawbacks and produced an ester
Talebian-Kiakalaieh and Amin 2015; Deb et al. 2017).
yield & conversion nearly equal to that produced in other

CONTACT Rakesh Kumar rkumar@rgipt.ac.in Department of Chemical Engineering, Rajiv Gandhi Institute of Petroleum Technology, Jais, Amethi 229304, India
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/ljge.
© 2018 Taylor & Francis Group, LLC
630 Z. HUSSAIN AND R. KUMAR
processes. For example, the product yield of ~94% was
achieved by Minami and Saka (2006) at 270°C & 200 bar
using 1:0.9 rape seed oil to methanol ratio (v/v). Melo-
Júnior et al. (2009) observed C18 fatty acids conversion of
60% non-catalytically in a short period of time (60 min) under
microwave irradiation. In 5 hrs reaction time at 1:4 Jatropha
oil to methanol ratio (w/v), 190°C and 27.1 bars ~95% con-
version was reported by Rani et al (2016). Cho et al. (2012)
achieved a desirable final product acid value of less than
0.5 mg KOH/g using palm fatty acid distillate and methanol
in 3 hrs at 290°C and 8.5 bar. Pinnarat and Savage (2010)
investigations revealed that the noncatalytic esterification can
be carried out smoothly even at subcritical conditions and
also reported the feasibility of sub critical esterification pro-
cess in tolerating moderate content of water in the feed oil
and its presence even enhances the effectiveness of subcritical
esterification process (Du et al. 2013).
Simulation study carried out by Haas et al. (2006), West,
Posarac, and Ellis (2008) and Lee, Posarac, and Ellis (2011) in
Aspen HYSYS demonstrated the continuous biodiesel produc-
tion processes using edible oils and assessed the technical and
economic feasibility. In view of above-revised literature, there
is a lag in kinetic modeling and process simulation studies of
non-catalytic esterification process.
The objective of the present work is to conduct the detailed
study of non-catalytic esterification of Karanja oil. The effect
of process variables such as speed of stirring, reaction time,
temperature, and ratio of oil & methanol was investigated in a
batch experiment. The reversible reaction was modeled and
the kinetic parameters were evaluated using experimental
data. Further, the obtained kinetic parameters were used to
simulate the process in Aspen plus®. Sensitivity analysis was
carried out to obtain the optimal process parameters along
with heat integration.
2. Materials and methods
2.1. Materials
Karanja oil of physicochemical properties provided by the
vendor given in Table 1 was purchased from Suyash herbs
exports Pvt. Ltd. Gujarat, India. The entire analytical grade
chemicals used in this work such as methanol, ethanol, potas-
sium hydroxide pellets, and phenolphthalein powder were
purchased from Sigma-Aldrich (India) Pvt. Ltd and used as
Table 1. Physico-chemical properties of Karanja oil.
Properties Value
Moisture (Dean & Stark method) 0.05%
Density@25°C 0.93 g/cc
Viscosity @40°C 40.07 c.st.
Color in inch cell on Lovibond scale 34.90
Impurities (insoluble in hexane) 0.43%
Acid value 62.80 mg KOH/g
Fatty acid composition (wt %) Palmitic acid 4.20
Stearic acid 2.90
Oleic acid 66.80
Linoleic acid 17.60
Arachidic acid 3.80
Behenic acid 4.70
obtained. Argon gas cylinder was purchased from Sigma gases
& services, New Delhi.
2.2. Sample analysis
Acid value and FFA content of the Karanja oil were deter-
mined using the titrimetric method and estimated using eq.1
and 2 as given below (“Manual of Methods Analysis of Foods:
Oils & Fats” 2012).
56:1  V  N
Acid value ¼ (1)
W
28:2  V  N
FFA as oleic acid ¼ (2)
W
where
V is the volume of standard potassium hydroxide used, ml;
N is the normality of standard potassium hydroxide solution;
W is the weight of the sample, gram.
2.3. Experimental setup
The esterification reaction was conducted in an experimental
setup as shown in Figure 1. The batch reactor having a
capacity of 25 ml was made up of stainless steel and placed
inside a jacketed vessel. The reactor was fitted with the pres-
sure gauge for measuring the reactor pressure. A magnetic
stirrer with the heater was used as primary source for heating
and mechanical agitation. There was a provision for the inert
gas inlet to the reactor for maintaining the constant pressure.
Also, the constant temperature inside the reactor was ensured
using a chiller.
2.4. Experimental procedure
The Karanja oil was washed thoroughly with hot deionized
water (~80°C) to remove soluble impurities present in the oil
and then it was treated with silica gel to remove the water
content. Further, the oil was heated at 105°C for 1 h to
remove the trace amount of water. Therefore, the water con-
tent present in the oil was assumed as negligible. In a typical
run, the reactor was charged with ~4.34 g oil and ~17.36 ml
methanol. These reactant amounts were selected so as to meet
the required criteria of volumetric filling fraction (ratio of the
volume of reactants charged to the total reactor volume, f). It
controls the phases that co-exists at reaction conditions.
Maintaining higher values of f under subcritical conditions
leads to liquid phase reactions when the reaction is to be
performed under inevitable subcritical conditions. Therefore,
we have selected the values of f such that the reactants remain
in homogeneous phase.
After charging the reactants into the reactor, the top of the
reactor was covered and fastened using bolts. It was then
placed into the jacketed vessel and the whole assembly was
kept on a magnetic stirrer equipped with a heater. The reac-
tion temperature and agitation speed were varied by the help
of magnetic stirrer and argon gas was supplied to maintain
the desired pressure inside the reactor. The variables affecting
the reaction such as agitation speed, temperature, reaction

Figure 1. Schematic of the experimental setup.
1- Reactor; 2- Outer shell/Jacket to reactor; 3- Magnetic stirrer with heater; 4- Magnetic bar; 5-Digital thermometer; 6, 9, 10- Pressure gauges; 7- Argon gas cylinder; 8-
Pressure regulator; 11-Gas inlet pipe to the reactor; 12-Chiller; 13-Chiller opening/knob for coolant feed; 14-Chilling temperature indicator or set point panel; 15-
Coolant inlet pipe to the reactor; 16-Coolant outlet pipe to the reactor.
time, oil to methanol ratio were studied. The reaction was
stopped at various time intervals and the reaction mass was
analyzed to estimate the acid value. The conversion of the
FFA was calculated using equation-3 (Vijaya Lakshmi,
Venkateshwar, and Satyavathi 2011; Abdala et al. 2014;
Endut et al. 2017);
ðAcid ValueÞt¼0  ðAcid ValueÞt¼t
Conversion ð%Þ ¼
ðAcid ValueÞt¼0
 100
3. Results and discussions
3.1. Feasibility of reaction at reactor conditions/phase
equilibrium
The reaction conditions employed to study the conversion of
FFA are above the boiling point of methanol and also below
its critical conditions (Tc = 239.35°C & Pc = 80.8 bar).
Therefore, it is usually expected the presence of liquid and
vapor phases. To maintain only one phase in the reactor the
volumetric filling fraction (ratio of the volume of reactants
charged to the total reactor volume, f) is varied to study the
phases coexisted at the reaction conditions. In addition to
experimental runs, we also performed phase equilibrium cal-
culations using ASPEN plus ver.8.6 software. ‘PRMHV2ʹ
method which uses Peng-Robinson equation of state (EOS)
with modified Huron-Vidal mixing rules were used to
INTERNATIONAL JOURNAL OF GREEN ENERGY 631
estimate thermodynamic properties of the present non-elec-
trolytic system. Flash2 block was used to study the phase
equilibrium of reactants. The input for the calculations was
the composition of feed streams, temperature & pressure or
vapor fraction. These provide the result for reactor pressure, a
fraction of coexisting phases when multiple phases were pre-
sent or pressure at which the reactor must be maintained
when a single phase is to exist. For instance, at a 1:5 ratio of
oil to methanol, 220°C and f = 0.88, the calculated pressure is
(3) 56 bar. Aspen plus reveals that at these conditions most of the
reactants are in liquid phase. Whereas, at reaction tempera-
ture of 220°C & f = 0.65 the calculated pressure is 10 bar.
According to aspen most of the reactants are in vapor phase at
these conditions. In the study carried out by Pinnarat &
Savage (2010), reported that, at 230°C & f = 0.56 the calcu-
lated pressure was 5.2 MPa observing a single liquid phase, at
250°C & f = 0.26 the calculated pressure was 5.28 MPa
observing 50/50 liquid & vapor and at 250°C with f = 0.04,
mostly vapors are observed using oleic acid & ethanol as
esterification reactants. In addition, they concluded that
maintaining low and high values of f leads to gas and liquid
phase systems, respectively, and concluded that supercritical
conditions are not at all required to obtain the desirable
esterification conversion. To evaluate the reaction progress
& kinetics of the reaction, a single phase is ensured by choos-
ing the reactor conditions in such a way that, at low tempera-
tures higher volumetric filling fraction is maintained and at
high temperatures lower oil to methanol ratios is maintained
632 Z. HUSSAIN AND R. KUMAR
coveting subcritical conditions. Additionally, optimal agita-
tion was ensured to suppress the heterogeneity of reactant
mixture.
3.2. Effect of agitation speed
To study the effect of agitation on the esterification reaction,
the speed was varied from 500 to 700 rpm. The agitation
speed has a strong influence on the conversion of the reaction
as shown in Figure 2. Due to the heterogeneous nature of the
reactants (oil and methanol), conversion of FFA was increas-
ing with increase in the agitation speed. It was observed that
the conversion of FFA at 500 rpm are much lower than the
conversions at 600 rpm. This behavior can be attributed to the
fact that there exists mass transfer resistance at lower agitation
which becomes negligible at a higher speed of agitation. This
fact has been reported by various researchers and showed that
the initial phase of the reaction is mass transfer controlled
(Noureddini and Zhu 1997; Alcantara et al. 2000; Ma,
Clements, and Hanna 1999; Skelland and Kanel 1990;
Skelland and Ramsay 1987). In the mixing of two immiscible
liquids such as oil and methanol there exist two distinct
phases called dispersed and continuous phases. The main
purpose of agitation is to completely incorporate the dis-
persed phase into the continuous phase by promoting good
contact between the phases so that the mass transfer rate
increases by extending interfacial area between them. The
speed at which the dispersed phase is completely incorporated
into the continuous phase is termed as minimum agitation
speed or critical speed (Vijaya Lakshmi, Venkateshwar, and
Satyavathi 2011). The conversion values become almost equal
on further increasing the agitation speed from 600 to 700
rpm. Therefore, 600 rpm was selected as the critical speed of
agitation and used for all the experiments.
1.0
0.8
Fractional conversion
0.6
0.4
0.2
0.0
0 2 4 6
Time, h
Figure 2. Effect of agitation on conversion at 220°C, 10 bar, and 1:6 oil to
methanol ratio (w/v).
3.3. Effect of reaction time & temperature
The present study reveals the increasing trend of FFA con-
version with time and reached 43% after 1 h of reaction time
as shown in Figure 3. A maximum conversion of 96% was
observed in 7 h at 220°C and 1:6 ratio of Karanja oil to
methanol (w/v). The reaction approached the equilibrium
condition after 7 h; therefore, no change in conversion was
observed. Due to the presence of distinct layer between the oil
and methanol, the reaction between them is not the sponta-
neous one. Therefore, change in the reaction temperature may
affect the reaction rate. To study the effect of temperature it
was varied in the range of 190°C to 220°C. The increase in
temperature resulted in the higher conversion values from
82% at 190°C to 96% at 220°C as shown in Figure 3. This
shows the endothermic nature of the reaction. These results
are close to the results of 95.1% equilibrium conversion at
190°C & 1:4 (w/v) Jatropha oil to methanol ratio in 5 h using
batch reactor (Rani et al. 2016) and 99.85% conversion at 290°
C in 3 h reaction time when 2.4 g/min methanol was supplied
to esterify 860 g of PFAD in a semi-batch reactor (Cho et al.
2012).
3.4. Effect of oil to methanol (w/v) ratio
The ratio of oil to methanol is an important parameter in both
esterification and transesterification reaction. Esterification of
FFA is the reversible reaction and oil to methanol ratio may
have the strong influence on the ester formation. Methanol
used in excess may help to shift the equilibrium towards
product formation. To study its effect, we performed the
experiments at three different oil/methanol ratios; 1:4, 1:5 &
1:6 as shown in Figure 4. It was observed that increase in the
oil to methanol ratio increases the conversion of FFA. For
instance, at 1:4 ratio only 79% conversion was achieved as
compared to 96% at 1:6 oil/methanol ratio (w/v). The conver-
sion values for 1:5 ratios were found almost equal to the
1.0
0.8
Fractional conversion
0.6
0.4
700 r.p.m 190oC
600 r.p.m
0.2 200oC
500 r.p.m
210oC
220oC
8 0.0
0 2 4 6 8 10 12 14
Time, h
Figure 3. Effect of temperature on conversion at 1:6 oil to methanol ratio (w/v).

1.0
0.8
Fractional conversion
0.6
0.4
0.2
0.0
0 1 2 3 4 5
Time, h
Figure 4. Effect of oil/methanol ratio on conversion at 220°C temperature.
values at 1:6 ratios; therefore 1:5 (w/v) oil/methanol ratios was
treated as optimal and used for further study. This result was
in good agreement with the similar study conducted for
Jatropha oil (Rani et al. 2016) where optimum oil/methanol
ratio was found to be 1:4(w/v). The deviation in oil/methanol
ratio may be attributed to higher acid value in the present
study, 62.8 mg KOH/g Karanja oil than the acid value of
54.43 mg KOH/g Jatropha oil.
3.5. Equilibrium constant & kinetic modeling
Equilibrium conditions are reached when the composition of
the reaction mixture does not change further over a course of
time. The composition of reactants and products at equili-
brium are required to determine equilibrium constant experi-
mentally. Principally rigorous thermodynamic methods based
on activity coefficients are the correct method to determine
the equilibrium constant as given in Equation-4.
aME aH2 O
K ¼ i ðai Þνi ¼ ¼ Kγ KC
aFFA aM
where ai is the activity of species i, Kγ = Πi(γi)νi, is activity
coefficient, KC = Πi(Ci)νi is the equilibrium constant based on
concentration (Neumann et al. 2016). Activity-based treat-
ment of equilibrium constant was used by many researchers
(Hernández-Montelongo, García-Sandoval, and Aguilar-
Garnica 2015; Keurentjes, Janssen, and Gorissen 1994;
Neumann et al. 2016; Rönnback et al. 1997) and observed
that the concentration based treatment of data fits better in
the model (Hassan and Vinjamur 2013; Keurentjes, Janssen,
and Gorissen 1994; Rönnback et al. 1997). For example,
Keurentjes, Janssen, and Gorissen (1994) used both activa-
tion-based and concentration based approaches to study the
kinetics of esterification of tartaric acid with ethanol and
reported that both the approaches predict the kinetic para-
meters reasonably, but prediction with later perform better.
INTERNATIONAL JOURNAL OF GREEN ENERGY 633
Similarly, in the case of esterification of acetic acid with
methanol carried out by Rönnback et al. (1997) showed no
significant improvement in kinetic parameters estimation
with the activation-based approach and concluded that the
model fit is better when the concentration-based approach is
applied. Additionally (Hassan and Vinjamur 2013) carried out
an esterification of oleic acid as FFA in sunflower oil and
corroborated the observation of Keurentjes, Janssen, and
Gorissen (1994) & Rönnback et al. (1997), that concentra-
tion-based approach in estimating equilibrium composition is
reasonable. Moreover, studies carried out by Liu et al. (2009)
on liquid-liquid immiscibility of (i) oleic acid+ methanol
+ water, (ii) ME+ methanol+ water, (iii) oleic acid + methanol
1:4 (Oil: methanol) + oil, (iv) ME + methanol + oil systems observed the oil
1:5 (Oil: methanol)
+ oleic acid & methanol + oleic acid as completely miscible
1:6 (Oil: methanol)
systems but methanol + oil is partially miscible system. They
observed that the miscibility (inter solubility) of oil and
6 7 8 methanol increases with temperature and also with an
increase in FFA or oleic acid content in the Jatropha oil. A
single homogeneous phase was found when there is 20 wt%
oleic acid content is present in Jatropha oil (Hassan and
Vinjamur 2013). Similar solubility was found in the work of
Batista et al. (1999) using canola oil+ oleic acid+ methanol.
This seems that the vegetable oil type has a little effect on
mutual solubilities with methanol. In the present study, the
content of FFA is 31.55 wt% of Karanja oil and temperature is
high enough. Together gaining confidence from the work of
Vijaya Lakshmi, Venkateshwar, and Satyavathi (2011), where
they demonstrated the critical speed of agitation at which
dispersed phase is completely incorporated into the continu-
ous phase and used the concentration-based approach to
study the mixing regime of Karanja oil and methanol.
Therefore, from the credence of above-reviewed literature, it
is expected that there exists a single homogeneous phase in
the reactor and we used the concentration-based approach to
evaluate the kinetics of esterification of FFA in Karanja oil.
The esterification of FFA with methanol is represented by a
reversible reaction as shown below;
kr
FFA þ M Ð MEþH2 O (5)
kf
where M is methanol; Me is methyl ester and H2O is water.
(4) The rate of above reaction is expressed as;
dCFFA
rFFA ¼  ¼ kf CFFA CM  kr CMe CH2 O (6)
dt
where CFFA; CM; CME & CH2 O are the concentrations of FFA,
methanol, methyl ester, and water, respectively, in mg KOH/g
oil; kf kr are the reaction rate constants for the forward and
backward reactions, respectively, in (mg KOH h/g oil)−1.
Following assumptions were considered for simplifying the
rate expression;
(1) Due to the presence of mixing regime, only the pre-
sence of one phase is assumed.
(2) Neglecting mass transfer effects, the reaction is
assumed to be kinetically controlled.
(3) Since methanol has been used in excess, therefore, for-
ward reaction (esterification of FFA) was considered as
634 Z. HUSSAIN AND R. KUMAR

0
pseudo-first order and the backward reaction (hydro- dX 0 k 1  Xe
lysis of ester) was taken as a bimolecular second order. ¼ kf ð1  XÞ  f CFFAo X2
dt CFFAo X2e
0
Following expressions were obtained from material balance; dX kf
2 
¼ 2 Xe  XX2e  X2 þ X2 Xe
CFFA ¼ CFFAo ð1  XÞ; dt Xe
ð 0 ð
CMe ¼ CH2 O ¼ CFFAo  CFFA ¼ CFFAo X dX kf

2  ¼ 2 dt
where CFFAo is the initial concentration of FFA. Xe  XX2e  X2 þ X2 Xe Xe
On applying assumptions and materials balance expres- Integration of the above equation gives;
sions in equation (6) we get;   0
ðXe  1ÞX  Xe kf t
dCFFA 0 ln ¼ ð2  Xe Þ (9)
)  ¼ kf CFFA  kr CMe CH2 O X  Xe Xe
dt
Simplifying the above equation for conversion of FFA as a
dX 0
CFFAo ¼ kf CFFAo ð1  XÞ  kr C2FFAo X2 function of time, we get;
dt  0  
kf tð2Xe Þ
dX 0 Xe exp 1
[ ¼ kf ð1  XÞ  kr CFFAo X2 (7) Xe
dt X¼ 0 (10)
k tð2X Þ
Also, we know that at equilibrium; exp f Xe e  ðXe  1Þ

dX Equation-(10) can also be converted in linear form y = mx.

¼ 0 X ¼ Xe h i
dt 0

where y ¼ ln ðXe 1ÞXX

kt
XXe
e
; x ¼ t slope; m ¼ Xfe ð2  Xe Þ
Therefore, equation (7) becomes as;
0 The kinetic model along with the reaction conditions pro-
0 ¼ kf ð1  Xe Þ  kr CFFAo X2e
posed by various researchers for esterification of FFA has
0
1 kf X2e been summarized in Table 2. The current model shows rela-
K¼ ¼ tively new treatment of reaction kinetics based on the assump-
CFFAo kr 1  Xe
tions which are associated with ultimate facts of the reaction.
0
k 1Xe The experimental conversion vs time data at each temperature
kr ¼ f (8) was plugged into the linear form of equation-(9) to evaluate x
CFFAo X2e
and y values. The obtained values were plotted for each
where K and Xe are the equilibrium constant and equilibrium temperature as shown in Figure 5. The proposed model fits
conversions, respectively. experimental data very well at each temperature with
On substituting equation (8) in equation (7) we get; R2 = 0.98. Further, the slope of each straight line was used

Table 2. Kinetic models.

Reactor type &
Reference Reactants optimum conditions Assumptions Model equation
" k0 tð2X Þ #
Present study Karanja oil & Batch reactor (1) Forward reaction is pseudo first order f
e
Xe
methanol T = 220°C, (2) Backward reaction is second order Xe exp 1
*Ratio = 1:5 (w/v) X¼ 0
k tð2Xe Þ
P = 10 bar f
Xe
Xe = 96.0% exp ðXe 1Þ
t=7h 
(Cho et al. 2012) PFAD & Semi batch reactor (1) Reversible reaction is neglected CFA;t 0
ln CFA;0 ¼ kf t Where CFA,t, CFA,0 is fatty acid conc.
methanol T = 290°C (2) Overall reaction follows first order
*Ratio = 1:1.58 (w/v) at t = t & t = 0, respectively.
P = 8.5 bar
Xe = 99.9%
t=3h h i
Xe expð ðX1e 1ÞtÞ 1
(Narayana Prasanna Jatropha oil & Batch reactor Both forward and backward reactions are 2k1 CAo

Rani et al. 2016) methanol T = 190°C considered as second order X¼ Where CAo is the initial
expð ðX1e 1ÞtÞ 2ðXe 1Þ
2k1 CAo
*Ratio = 1:4 (w/v)
P = 27.1 bar acid value
Xe = 95.1% k1 is the forward reaction rate constant
t=5h
(Abdala et al. 2014) Oleic acid & Plug flow reactor No kinetic study -
ethanol T = 300°C
Molar ratio = 1:6
P = 200 bar
Xe = 88.0%
t = 0.33 h
*Ratio represents oil: methanol
 INTERNATIONAL JOURNAL OF GREEN ENERGY 635

5 1.0

4 0.8
ln[((Xe-1)X-Xe)/(X-Xe)]

Fractional conversion
0.6
3

0.4
2
220oC
o
220 C 210oC
0.2
1 o
210 C 200oC
o
200 C 190oC
o
190 C 0.0
0
0 2 4 6 8 10 12 0 2 4 6 8 10 12 14

Time, h time, h

Figure 5. Validation of the kinetic model. Figure 6. Experimental and model predicted conversion at various
temperatures.

to calculate the forward reaction rate constant (kf). The calcu-

lated kf values were used in equation-(8) to obtain the back-
ward reaction rate constant. The obtained reaction parameters
have been presented in Table 3. Various researchers have used
trial and error method to calculate equilibrium conversion
and rate constants, in the case of limited experimental data
and unavailability of experimental equilibrium conversion
(Rani et al. 2016; Pinnarat and Savage 2010; Alenezi et al.
2010). To check the model applicability for such conditions,
conversion vs time data was regressed using Levenberg-
Marquardt algorithm. The regression of experimental data
was carried out by providing initial guess values of unknown
parameters Xe and kf until equation-10 fits well. The model
predicts the conversion vs time profile closely matching
(R2 = 0.99) with experiments at each temperature as shown
in Figure 6. The parameters such as rate constants and equili-
brium conversion obtained through trial and error method
using equation-10 have been reported in Table 3. The para-
meters match well with the experimental equilibrium conver-
sion as well as rate constant at each temperature.
3.6. Effect of temperature on reaction rate constant
The effect of temperature on reaction rate constant was stu-
died using Arrhenius equation as given below;
ΔE
lnk ¼  þ lnko
RT
lnk vs 1/T data was plotted for both rate constants (kf, kr) as
shown in Figure 7. Activation energies (ΔE) and frequency
factors (ko) for the forward and backward reactions were
evaluated from the slope and intercept of the straight lines,
respectively, and tabulated in Table 4.
3.7. Effect of temperature on equilibrium constant
The effect of temperature on the equilibrium constant was
studied using Vant Hoff’s equation. The lnK vs inverse of
temperature data was fitted to a straight line as shown in
Figure 8. The slope and intercept of the line were used to
calculate enthalpy (ΔH) & entropy (ΔS) of the reaction,
respectively. The values of and ΔH & ΔS are +29.79 kJ/mol
and +0.121 kJ/mol.K, respectively, as given in Table 4. The
positive sign of ΔH indicates the endothermic nature of the
reaction and + ΔS shows the increased entropy of the system.
To find whether the reaction is spontaneous or not, the Gibbs
free energy was also calculated at each temperature using the
following Equation-12;
ΔH ΔS
þ lnK ¼  (12)
RT R
The calculated Gibb’s free energy change (ΔG) was found to
be negative at each temperature as shown in Table 3. The
large negative values of ΔG along with + ΔH and + ΔS con-
firms that the reaction is spontaneous at high temperatures.
(11)
Experimentally observed conversion values were plotted

Table 3. Calculated reaction parameters.

Experimental Trial and error method
T°C Xe kf’ (min−1) kr (g/mgKOH min) R2 Xe kf’ (min−1) kr (g/mgKOH min) R2
190 0.941 0.295 3.09E-4 0.991 0.964 0.291 1.79E-4 0.999
200 0.950 0.372 3.31E-4 0.997 0.955 0.368 2.89E-4 0.999
210 0.956 0.469 3.60E-4 0.995 0.966 0.461 2.67E-4 0.999
220 0.962 0.641 4.18E-4 0.987 0.972 0.638 3.01E-4 0.999
636 Z. HUSSAIN AND R. KUMAR

0 1.0

-1
+5%
0.8
-2

Predicted fractional conversion

-3
0.6 -5%
-4
lnkf'
lnk

lnkr
-5
0.4
o
-6 190 C
o
200 C
-7 0.2 o
210 C
o
-8 220 C

0.00200 0.00204 0.00208 0.00212 0.00216 0.0

0.0 0.2 0.4 0.6 0.8 1.0
1/T, K-1
Observed fractional conversion
Figure 7. Arrhenius plot.
Figure 9. Comparison of experimental and model predicted conversions.

Table 4. Calculated Arrhenius and thermodynamic parameters.

Parameters Forward reaction Backward reaction 4. Process modeling and simulation
#
ko 86249.54* 0.04 4.1. Modeling
ΔE 48.53 18.74
ΔG −26.42 −27.64 Process simulation is known to provide valuable information
ΔH 29.79
ΔS 0.12 about the operating conditions efficiently in a short period of
Units: *min−1; #g/mgKOH min; ΔE, ΔH & ΔG in kJ/mol; ΔS in kJ/mol °C time. In spite of the fact that there are some differences
between real process and process simulation, yet, they are
widely used to provide reliable information due to their
7.4
extensive component library, comprehensive thermodynamic
packages, and diverse computational methods (West, Posarac,
and Ellis 2008). The non-catalytic esterification process was
7.3 modeled using Aspen plus® simulator (ver.8.6). The complete
flow sheet of the process is shown in Figure 10. Karanja oil
was pumped using a feed pump (P-100) and resulting stream
7.2 (E-101) was introduced into the feed mixer (M-101). The
recycled methanol (E-108) was mixed with makeup methanol
7.1
in a mixer (M-100) and resulting methanol stream (E-102)
lnK

was sent to feed mixer (M-101). After mixing, reactants were

pumped again using mixed feed pump (P-101) and resulting
7.0 stream (E-103) was sent to a heat exchanger (HE-100) on the
tube side. The heated feed from the heat exchanger (E-104)
having zero vapor fraction enters the continuous stirred tank
6.9 reactor (R-100) where the esterification reaction takes place.
The product stream (E-105) of the reactor is passed through
6.8
the flash column (F-100) where pure methyl ester was col-
0.00200 0.00204 0.00208 0.00212 0.00216 lected as bottom product and unreacted methanol, as well as
water, was removed from the top as mixed stream (E-106). In
1/T, K-1
order to utilize the heat of mixed stream, it is passed through
Figure 8. Van’t Hoff plot. the shell side of the heat exchanger to heat the feed stream.
Product stream resulting from the heat exchanger (E-107) was
fed to the distillation column (D-100) in which pure methanol
against the model predicted conversions to test the statistical was recovered from top and recycled (E-108) to the mixer (M-
significance of the data using Student’s paired t-test. It can be 100). The by-product water was recovered from the bottom of
seen from Figure 9 that the proposed model satisfactorily the distillation column. The CSTR, flash column, distillation
represents all the data.
 INTERNATIONAL JOURNAL OF GREEN ENERGY 637

Figure 10. Flow diagram of esterification process.

Nomenclature:M-100, 101:- Feed mixers; P-100, 101:- Feed pumps; HE-100:- Heat exchanger; F-100:- Flash column; D-100:- Distillation column; R-100:- Continuous
stirred tank reactor; QH: - Heat exchanger heat duty; QR: - Reactor duty; QF1:- Flash column duty; QC: - Condenser duty; QRe: - Reboiler duty.

column shown in the flow sheet were modeled using RCSTR,
FLASH-2, and RADFRAC subroutines of Aspen software,
respectively.
4.2. Simulation
The simulation was carried out by importing the required
databanks and specifying all the components present in the
process. Activity coefficient based NRTL thermodynamic
model was selected for estimation of properties as well as
the calculation of phase equilibrium. Triglycerides (TG) are
assumed to be homogeneous and each TG composed of only
one type of fatty acid radical. The flow rate of Karanja oil in
feed stream (E-101) was selected to be 1050 kg/h with com-
position as given in Table 5. In order to lower the process
complexity and due to the high oleic acid composition in the
present Karanja oil, the FFA was assumed entirely as oleic
acid. Methanol was fed continuously to the reactor
(4547.51 m3) at the rate of 6300 l/h, i.e. at 1:6 (w/v) ratio.
The kinetic parameters obtained through experiment such as
ko = 86249.54 min−1, ΔE = 48.534 kJ/mol for the forward
reaction and ko = 0.039 (g)/(mgKOH min), ΔE = 18.744 kJ/
mol for the backward reaction were used in RCSTR subrou-
tine of the reactor model. The reaction product of oleic acid
(FFA) and methanol were designated as methyl-oleate
(methylo). After completion of the reaction, the product was
separated using a flash drum. The temperature and pressure
of the flash drum were maintained at 220°C and 10 bar,
respectively. The pure product was collected from the bottom
and water as well as methanol was collected from the top.
Product recovery greater than 87% (ratio of molar flow rates
of all the components excluding water & methanol in methyl

Table 5. Stream results of simulated flow sheet.

E-101 E-102 E-103 E-104 E-105 ME E-106 E-107 E-108 Water Make-up
T 25 63.89 59.72 64.52 220 220 220 210.5 64.54 75.67 25
(°C)
P 1 1 1 1 10 10 10 10 1 1 1
(bar)
Mass flow rate 1050 5067.07 6117.07 6117.07 6117.07 970.44 5146.63 5146.63 4976.96 169.67 90.11
(kg/h)
Molar flow rate 2.02 158.14 160.16 160.16 160.16 2.13 158.03 158.03 155.33 2.70 2.81
(kmol/h)
Molar composition OOO 0.217 0 2.75E-3 2.75E-3 2.75E-3 0.179 3.53E-4 3.53E-4 0 0.020 0
O 0.607 0 7.68E-3 7.68E-3 1.14E-5 7.87E-4 0 0 0 0 0
Methanol 0 0.99 0.98 0.98 0.98 0.17 0.99 0.99 0.99 0.45 1
Water 0 1.11E-4 1.09E-4 1.09E-4 7.78E-3 3.03E-3 7.85E-3 7.85E-3 1.13E-4 0.452 0
M-O 0 0 0 0 7.67E-3 0.481 1.26E-3 1.26E-3 0 0.073 0
PPP 0.028 0 3.58E-4 3.58E-4 3.58E-4 0.026 0 0 0 0 0
SSS 0.017 0 2.24E-4 2.24E-4 2.24E-4 0.016 0 0 0 0 0
LLL 0.108 0 1.38E-3 1.38E-3 1.38E-3 0.103 0 0 0 0 0
AAA 0.020 0 2.57E-4 2.57E-4 2.57E-4 0.019 0 0 0 0 0
ME: Methyl Ester, OOO: triolein, O: oleic acid, M-O: methyl-oleate, PPP: tripalmitin, SSS: tristearin, LLL: trilenolein,
AAA: triarachidic.
638 Z. HUSSAIN AND R. KUMAR

ester stream to the molar flow rate of all the components to E-106 stream as shown in Figures 11&12. Whereas, when
including methanol and water in E-105 stream) was observed the flash column is operated at 10 bar pressure, a good
with the minor loss of methyl-oleate and triolein into the top esterified product recovery was observed at the expense of
product. The top product was then passed through the heat allowing the little amount of water (0.0065 kmol/h) and
exchanger to recover heat followed by separation of water and methanol (0.37 kmol/h) into the ester stream. Therefore,
recovery of methanol in a distillation column. Distillation the flash column pressure was selected as 10 bar. The
column was modeled using rigorous Radfrac subroutine of complete flow sheet simulation results are presented in
the Aspen simulator. The required inputs for Radfrac were Table 5. The calculated heat duties of each unit are as
calculated first by DSTWU model. The molar flow rate, com- follows:
position, and stream conditions of E107 were given as input
to the DSTWU unit. The light key (methanol) and heavy key
(water) recoveries were specified in DSTWU as 0.99 and 0.01,
respectively. DSTWU simulation result shows that at reflux
156.90
ratio of 1.3, 23 stages are required to get desired product
purity with 155.333 kmol/h distillate flow rate and suggested
156.85
to locate the feed on the 20th stage. This information was
used as input to the Radfrac and nearly pure (0.999) methanol

Methanol in E-106, kmolh-1

156.80
was recovered. The recovered methanol was recycled back to
the mixer-1 where it was mixed with the makeup methanol 156.75
stream.
156.70

156.65
4.3. Sensitivity analysis
To study the effect of flash column pressure on product 156.60
recovery, sensitivity analysis tool of Aspen plus was used to
determine the optimum pressure that would maximize the 156.55
molar flow of all the components other than water &
156.50
methanol in the ester stream. Sensitivity results show that 2 4 6 8 10
when the flash column is operated at 2 bar, the maximum
Pressure, bar
amount of methanol (156.87 kmol/h) and water (1.2430
kmol/h) are recovered in methanol + water (E-106) stream Figure 11. Effect of flash column operating pressure on methanol mole flow in
but there is a considerable loss of triolein and methyl oleate E-106 stream.

0.8
Triolein
0.7 Methyloleate
M olar flow in E-106, kmolh-1

0.6

0.5

0.4

0.3

0.2

0.1

0.0
2 4 6 8 10
Pressure, bar
Figure 12. Effect of flash column operating pressure on triolein and methyl-oleate mole flow in E-106 stream.

Heat exchanger heat duty (QH) = 26 kW; Reactor duty
(QR) = 2038 kW; Flash column duty (QF1) = 0; Condenser
duty (QC) = −3481 kW; Reboiler duty (QRe) = 1601 kW.
The simulation results showed the conversion of oleic acid
to be 99.85% which was close to the experimental conversion
(96.21%). The little deviation in the predicted conversion may
be attributed to the fact that in simulation only oleic acid was
used to represent FFA. However, in experiments, FFA was a
complex blend of various fatty acids in Karanja oil.
5. Conclusions
The experimental result shows that the 96.21% FFA conversion
can be achieved in 7 h of reaction time at 220°C with 1:6 (w/v)
oil/methanol ratios. The kinetics of the reaction was well mod-
eled using the pseudo first order in the forward direction and
second order bimolecular type reaction in the reversible direc-
tion. The calculated kinetic parameters were found as kf’ (min-
−1
) = 0.0106, kr (g/mgKOH min) = 6.96E-6. The activation
energy for the forward and backward reactions was 48.534
and 18.744 kJ/mol, respectively. The values of enthalpy
(ΔH > 0), entropy (ΔS > 0) and Gibbs free energy (ΔG > 0)
confirm that the reaction is spontaneous only at high tempera-
tures subject to TΔS > ΔH. Simulation study revealed that
99.85% conversion of oleic acid may be achieved.
Acknowledgments
We would like to acknowledge the Rajiv Gandhi Institute of Petroleum
Technology (RGIPT), Jais, India for providing the facilities & financial
support for this work.
ORCID
Zakir Hussain http://orcid.org/0000-0002-2213-0573
Rakesh Kumar http://orcid.org/0000-0001-9926-6156
References
Abdala, A., C. De Araujo, V. A. Dos Santos Garcia, C. P. Trentini, L. C.
Filho, E. A. Da Silva, and C. Da Silva. 2014. Continuous catalyst-free
esterification of oleic acid in compressed ethanol. International
Journal of Chemical Engineering. Article ID 803783. doi:10.1155/
2014/803783.
Alcantara, R., J. Amores, L. Canoira, E. Fidalgo, M. J. Franco, and A.
Navarro. 2000. Catalytic production of biodiesel from soy-bean oil,
used frying oil and tallow. Biomass and Bioenergy 18 (6):515–27. doi:
10.1016/S0961-9534(00)00014-3.
Alenezi, R., G. A. Leeke, J. M. Winterbottom, R. C. D. Santos, and A. R.
Khan. 2010. Esterification kinetics of free fatty acids with supercritical
methanol for biodiesel production. Energy Conversion and
Management 51 (5):1055–59. doi:10.1016/j.enconman.2009.12.009.
Alper Tapan, N., R. Yıldırım, and M. Erdem Günay. 2016. Analysis of
past experimental data in literature to determine conditions for high
performance in biodiesel production. Biofuels, Bioproducts and
Biorefining 10 (4):422–34. doi:10.1002/bbb.2016.10.issue-4.
Anitescu, G., A. Deshpande, and L. L. Tavlarides. 2008. Integrated
technology for supercritical biodiesel production and power cogenera-
tion integrated technology for supercritical biodiesel production and
power cogeneration. Energy 12:1391–99.
Axelsson, L., M. Franzén, M. Ostwald, G. Göran Berndes, G. Lakshmi,
and N. H. Ravindranath. 2012. Perspective: Jatropha cultivation in
INTERNATIONAL JOURNAL OF GREEN ENERGY 639
Southern India: Assessing farmers’ experiences. Biofuels, Bioproducts
and Biorefining 6 (3):246–56. doi:10.1002/bbb.1324.
Bassan, I. A. L., D. R. Nascimento, R. A. S. San Gil, M. I. P. Da Silva, C. R.
Moreira, W. A. Gonzalez, A. C. Faro, T. Onfroy, and E. R. Lachter. 2013.
Esterification of fatty acids with alcohols over niobium phosphate. Fuel
Processing Technology 106:619–24. doi:10.1016/j.fuproc.2012.09.054.
Batista, E., S. Monnerat, K. Kato, L. Stragevitch, and A. J. A. Meirelles.
1999. Liquid-liquid equilibrium for systems of canola oil, oleic acid,
and short-chain alcohols. Journal of Chemical and Engineering Data
44 (6):1360–64. doi:10.1021/je990015g.
Berrios, M., M. A. Martín, A. F. Chica, and A. Martín. 2010. Study of
esterification and transesterification in biodiesel production from used
frying oils in a closed system. Chemical Engineering Journal 160
(2):473–79. doi:10.1016/j.cej.2010.03.050.
Chai, M., T. Qingshi, L. Mingming, and Y. Jeffrey Yang. 2014.
Esterification pretreatment of free fatty acid in biodiesel production,
from laboratory to industry. Fuel Processing Technology 125:106–13.
doi:10.1016/j.fuproc.2014.03.025.
Cho, H. J., S. H. Kim, S. W. Hong, and Y. K. Yeo. 2012. A single step
non-catalytic esterification of Palm Fatty Acid Distillate (PFAD) for
biodiesel production. Fuel 93:373–80. doi:10.1016/j.fuel.2011.08.063.
Deb, A., J. Ferdous, K. Ferdous, M. R. Uddin, M. R. Khan, and M. W.
Rahman. 2017. Prospect of castor oil biodiesel in Bangladesh: Process
development and optimization study. Internatuional Journal of Green
Energy 14 (12):1063–72. doi:10.1080/15435075.2017.1357558.
Du, Z., Z. Tang, H. Wang, J. Zeng, Y. Chen, and E. Min. 2013. Research
and development of a sub-critical methanol alcoholysis process for
producing biodiesel using waste oils and fats. Chinese Journal of
Catalysis 34 (1):101–15. doi:10.1016/S1872-2067(11)60490-7.
Endut, A., S. H. Y. S. Abdullah, N. H. M. Hanapi, S. H. A. Hamid, F.
Lananan, M. K. A. Kamarudin, R. Umar, H. Juahir, and H. Khatoon.
2017. Optimization of biodiesel production by solid acid catalyst
derived from coconut shell via response surface methodology.
International Biodeterioration & Biodegradation 124:250–57.
doi:10.1016/j.ibiod.2017.06.008.
Gandhi, B. S., and D. S. Kumaran. 2014. The production and optimiza-
tion of biodiesel from crude jatropha curcas oil by a two step process
—an indian case study using response surface methodology.
International Journal of Green Energy 11 (10):1084–96. doi:10.1080/
15435075.2013.829777.
Gómez-Castro, F., V. Israel, J. Rico-Ramírez, G. Segovia-Hernández, and
H.-C. Salvador. 2011. Esterification of fatty acids in a thermally
coupled reactive distillation column by the two-step supercritical
methanol method. Chemical Engineering Research and Design 89
(4):480–90. doi:10.1016/j.cherd.2010.08.009.
Gupta, A. K., M. R. Kiro, and G. Deo. 2016. Biodiesel production from a
free fatty acid containing Karanja oil by a single-step heterogeneously
catalyzed process. International Journal of Green Energy 13 (5):489–
96. doi:10.1080/15435075.2014.977440.
Haas, M. J., A. J. McAloon, W. C. Yee, and T. A. Foglia. 2006. A process
model to estimate biodiesel production costs. Bioresource Technology
97 (4):671–78. doi:10.1016/j.biortech.2005.06.006.
Hassan, S. Z., and M. Vinjamur. 2013. Analysis of sensitivity of equili-
brium constant to reaction conditions for esterification of fatty acids
with alcohols. Industrial and Engineering Chemistry Research 52
(3):1205–15. doi:10.1021/ie301881g.
Hernández-Montelongo, R., J. P. García-Sandoval, and E. Aguilar-
Garnica. 2015. On the non-ideal behavior of the homogeneous ester-
ification reaction: A kinetic model based on activity coefficients.
Reaction Kinetics, Mechanisms and Catalysis 115 (2):401–19.
doi:10.1007/s11144-015-0848-x.
Hussain, Z., B. H. Mohammad, and R. Kumar. 2016. UsageSpecific
biodiesel production with and without catalytic booster. Materials
Today: Proceedings 3:4115–20. doi:10.1016/j.matpr.2016.11.083.
Keurentjes, J. T. F., G. H. R. Janssen, and J. J. Gorissen. 1994. The
esterification of tartaric acid with ethanol: kinetics and shifting the
equilibrium by means of pervaporation. Chemical Engineering Science
49 (24):4681–89. doi: 10.1016/S0009-2509(05)80051-X.
Kolyaei, M., G. Zahedi, M. M. Nasef, and A. Azarpour. 2016.
Optimization of biodiesel production from waste cooking oil using
640 Z. HUSSAIN AND R. KUMAR
ion-exchange resins. International Journal of Green Energy 13 (1):28–
33. doi:10.1080/15435075.2014.909354.
Lee, A. F., J. A. Bennett, J. C. Manayil, and K. Wilson. 2014.
Heterogeneous catalysis for sustainable biodiesel production via ester-
ification and transesterification. Chemical Society Reviews 43:7887–
916. doi:10.1039/C4CS00189C.
Lee, S., D. Posarac, and N. Ellis. 2011. Process simulation and economic
analysis of biodiesel production processes using fresh and waste vege-
table oil and supercritical methanol. Chemical Engineering Research
and Design 89 (12):2626–42. doi:10.1016/j.cherd.2011.05.011.
Liu, Y. ;., H. Lu, C. Liu, and B. Liang. 2009. Solubility measurement for
the reaction systems in pre- esterification of high acid value Jatropha
Curcas L. oil. Journal of Chem. Eng 54:1421–25.
Long, Y.-D., and Z. Fang. 2012. Hydrothermal conversion of glycerol to
chemicals and hydrogen: Review and perspective. Biofuels,
Bioproducts and Biorefining 6 (6):686–702. doi:10.1002/bbb.1345.
Lou, W. Y., Q. Guo, W. J. Chen, M. H. Zong, W. Hong, and T. J. Smith.
2012. A highly active bagasse-derived solid acid catalyst with proper-
ties suitable for production of biodiesel. ChemSusChem 5 (8):1533–41.
doi:10.1002/cssc.201100811.
Ma, F. L., D. Clements, and M. A. Hanna. 1999. The effect of mixing on
transesterification of beef tallow. Bioresource Technology 69 (3):289–
93. doi:10.1016/S0960-8524(98)00184-9.
Manual of Methods Analysis of Foods: Oils & Fats. 2012. Food safety &
standards authority of India (Ministry of Health & Family Welfare)
Government of India, New Delhi. Lab Manual (2):24–26.
Melo-Júnior, C. A. R., C. E. R. Albuquerque, M. Fortuny, C. Dariva, S.
Egues, A. F. Santos, and A. L. D. Ramos. 2009. Use of microwave
irradiation in the noncatalytic esterification of C18 fatty acids. Energy
and Fuels 23 (1):580–85. doi:10.1021/ef800766x.
Minami, E., and S. Saka. 2006. Kinetics of hydrolysis and methyl ester-
ification for biodiesel production in two-step supercritical methanol
process. Fuel 85 (17–18):2479–83. doi:10.1016/j.fuel.2006.04.017.
Neumann, K., K. Werth, A. Martín, and A. Gorak. 2016. Biodiesel produc-
tion from waste cooking oils through esterification: Catalyst screening,
chemical equilibrium and reaction kinetics. Chemical Engineering
Research and Design 107:52–62. doi:10.1016/j.cherd.2015.11.008.
Noureddini, H., and D. Zhu. 1997. Kinetics of transesterification of
soybean oil. Journal of the American Oil Chemists’ Society 74
(11):1457–63. doi:10.1007/s11746-997-0254-2.
Pinnarat, T., and P. E. Savage. 2010. Noncatalytic esterification of oleic
acid in ethanol. Journal of Supercritical Fluids 53:53–59. doi:10.1016/j.
supflu.2010.02.008.
Pisarello, M. L., B. Dalla Costa, G. Mendow, and C. A. Querini. 2010.
Esterification with ethanol to produce biodiesel from high acidity raw
materials: kinetic studies and analysis of secondary reactions. Fuel
Processing Technology 91 (9):1005–14. doi:10.1016/j.fuproc.2010.03.001.
Quesada-Medina, J., and P. Olivares-Carrillo. 2011. Evidence of thermal
decomposition of fatty acid methyl esters during the synthesis of
biodiesel with supercritical methanol. Journal of Supercritical Fluids
56 (1):56–63. doi:10.1016/j.supflu.2010.11.016.
Rani, K. N. P., T. S. V. R. Neeharika, P. K. Thella, B. Satyavathi, and S.
Chintha. 2016. Kinetics of non-catalytic esterification of free fatty
acids present in Jatropha oil. Journal of Oleo Science 65 (5):441–45.
doi:10.5650/jos.ess15228.
Rathore, V., S. Tyagi, B. Newalkar, and R. P. Badoni. 2015. Jatropha and
Karanja oil derived DMC-biodiesel synthesis: A kinetics study. Fuel
140:597–608. doi:10.1016/j.fuel.2014.10.003.
Rönnback, R., T. Salmi, A. Vuori, H. Haario, J. Lehtonen, A. Sundqvist,
and E. Tirronen. 1997. Development of a kinetic model for the
esterification of acetic acid with methanol in the presence of a homo-
geneous acid catalyst. Chemical Engineering Science 52 (19):3369–81.
doi:10.1016/S0009-2509(97)00139-5.
Sara, M., S. K. Brar, and J. F. Blais. 2016. Comparative study between
microwave and ultrasonication aided in situ transesterification of
microbial lipids. RSC Adv 6 (61):56009–17. doi:10.1039/C6RA10379K.
Sebastian, J., C. Muraleedharan, and A. Santhiagu. 2017. Enzyme cata-
lyzed biodiesel production from rubber seed oil containing high free
fatty acid. International Journal of Green Energy 14 (8):687–93.
doi:10.1080/15435075.2017.1318754.
Shin, H. Y., S. H. Lee, J. H. Ryu, and S. Y. Bae. 2012. Biodiesel production
from waste lard using supercritical methanol. Journal of Supercritical
Fluids 61:134–38.
Shin, H. Y., S. M. Lim, S. Y. Bae, and S. C. Oh. 2011. Thermal decom-
position and stability of fatty acid methyl esters in supercritical
methanol. Journal of Analytical and Applied Pyrolysis 92 (2):332–38.
doi:10.1016/j.jaap.2011.07.003.
Skelland, A. H. P., and G. G. Ramsay. 1987. Minimum agitator speeds for
complete liquid-liquid dispersion. Industrial & Engineering Chemistry
Research 26 (1):77–81. doi:10.1021/ie00061a014.
Skelland, A. H. P., and J. S. Kanel. 1990. Minimum impeller speeds for
complete dispersion of non-newtonian liquid-liquid systems in baffled
vessels. Industrial & Engineering Chemistry Research 29 (7):1300–06.
doi:10.1021/ie00103a032.
Stacy, C. J., C. A. Melick, and R. A. Cairncross. 2014. Esterification of
free fatty acids to fatty acid alkyl esters in a bubble column reactor for
use as biodiesel. Fuel Processing Technology 124:70–77. doi:10.1016/j.
fuproc.2014.02.003.
Talebian-Kiakalaieh, A., and N. A. S. Amin. 2015. Single and two-step
homogeneous catalyzed transesterification of waste cooking oil:
Optimization by response surface methodology. International
Journal of Green Energy 12 (9):888–99. doi:10.1080/15435075.2014.
884501.
Torres, E. A., G. S. Cerqueira, T. M. Ferrer, C. M. Quintella, M. Raboni,
V. Torretta, and G. Urbini. 2013. Recovery of different waste vegetable
oils for biodiesel production: A pilot experience in Bahia state, Brazil.
Waste Management 33 (12):2670–74. doi:10.1016/j.wasman.2013.
07.030.
Van, T., D. Son, N. P. Trung, V. N. Anh, and H. N. Lan. 2016.
Optimization of esterification of fatty acid rubber seed oil for methyl
ester synthesis in a plug flow reactor. International Journal of Green
Energy 13 (7):720–29. doi:10.1080/15435075.2014.966372.
Vieitez, I., C. Silva, I. Alckmin, G. R. Borges, F. C. Corazza, J. Vladimir
Oliveira, and M. A. Grompone. 2009. Effect of temperature on the
continuous synthesis of soybean esters under supercritical ethanol.
Energy & Fuels 23 (12):558–63. doi:10.1021/ef800640t.
Vijaya Lakshmi, C. K., V. S. Venkateshwar, and B. Satyavathi. 2011.
Mixing characteristics of the oil-methanol system in the production
of biodiesel using edible and non-edible oils. Fuel Processing
Technology 92 (8):1411–17. doi:10.1016/j.fuproc.2011.01.015.
West, A. H., D. Posarac, and N. Ellis. 2008. Assessment of four biodiesel
production processes using HYSYS.plant. Bioresource Technology 99
(14):6587–601. doi:10.1016/j.biortech.2007.11.046.
Zong, M.-H., Z.-Q. Duan, W.-Y. Lou, T. J. Smith, and W. Hong. 2007.
Preparation of a sugar catalyst and its use for highly efficient produc-
tion of biodiesel. Green Chemistry 9 (5):434–37. doi:10.1039/b615447f.