You are on page 1of 7

Energy Conversion and Management 79 (2014) 599–605

Contents lists available at ScienceDirect

Energy Conversion and Management


journal homepage: www.elsevier.com/locate/enconman

Optimization of biodiesel production from soybean oil in a microreactor


Masoud Rahimi a,b,⇑, Babak Aghel a, Mohammad Alitabar b, Arash Sepahvand b, Hamid Reza Ghasempour b
a
CFD Research Center, Chemical Engineering Department, Razi University, Kermanshah, Iran
b
Department of Biotechnology–Chemical Engineering, Science and Research Branch, Islamic Azad University, Kermanshah, Iran

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

Article history: Transesterification of soybean oil with methanol in the presence of potassium hydroxide, as a catalyst, in
Received 14 October 2013 a microreactor has been investigated. The transesterification reaction was performed at specific condition
Accepted 30 December 2013 in circular tubes with hydraulic diameter of (0.8 mm). In order to further improve the biodiesel produc-
Available online 30 January 2014
tion, the experimental design was performed using Box–Behnken method. The results were analyzed
using response surface methodology. The influence of reaction variables including; molar ratio of meth-
Keywords: anol to oil (6:1–12:1), temperature (55–65 °C) and catalyst concentration (0.6–1.8 wt.%) and residence
Microreactor
time (20–180 s) under various flow rates of reactants (1–11 ml min1) on Fatty Acid Methyl Ester (FAME)
Biodiesel
Soybean oil
transesterification reaction was studied. The optimum condition was found at molar ratio of methanol to
Optimization oil (9:1), catalyst concentration (1.2 wt.%) and temperature (60 °C) with a FAME % of about 89%. Consid-
Response surface methodology ering optimum parameters, by changing the reactant residence time the FAME % was reached to 98% at
180 s.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction research focused on reducing process time and promote the effi-
ciency of the biodiesel process [13,14]. Hanh et al. [15] investigated
The limited fossil fuel resources along with the need of reduce transesterification of methyl esters in presence of 40 kHz ultra-
green house gases emissions were major impulses to the develop- sonic irradiation at room temperature. Microwave method to pro-
ment of alternative fuels. As a result, increased attention has been vide biodiesel using action ion-exchange resin particles (CERP)/PES
given to biofuels, such as biodiesel, which can be used as an alter- catalytic was assisted by Zhang et al. [16]. They showed that at the
native fuel in compression–ignition engines [1]. The other advan- optimal conditions, this method is a fast and easy way for
tage of biofuels is its contribution to reduce of CO2 emissions; producing biodiesel with a reaction time of 90 min. Among other
because of it comprises a closed carbon cycle [2,3]. Biodiesel pro- methods, in a study done by Yin et al. [17], supercritical transeste-
duced by transesterification of renewable resources, such as vege- rification of soybean oil with co-solvent attained methyl ester yield
table oils and animal fats with monohydric alcohol, which is of more than 98% during 10 min at temperature of 350 °C. They
commonly methanol, makes it biodegradable and nontoxic [4]. found that by adding hexane as co-solvent to carbon dioxide at
The transesterification reaction is three-step reversible reaction 300 °C, the yield of methyl ester was notably increased. From the
for converting the triglyceride into a mixture of esters (biodiesel) literature, continuous transesterification of oils with catalyst to im-
and glycerol in the presence of a catalyst. During transesterification prove the efficiency were proposed by many researchers [18–20].
reaction, the triglyceride is converted to the diglyceride, monoglyc- It is known that the main drawback of base-catalyzed biodiesel
eride and glycerol in the three steps, and one mol of ester produces process is high cost and time consumed during transesterification
at each step. and disposal of large amount of waste water. The cost of operation
Two types of catalysts including; homogeneous (acid or base) will be decreased if the time of process decreases. In order to in-
and heterogeneous (acid, base, and enzymatic) were used for crease the conversion efficiency of biodiesel under continuous con-
transesterification of vegetable oils [5–8]. Biodiesel has some dition, microreactor can be used. Microchannel reactor as a
disadvantages such as high feedstock cost, energy requirements, continuous reactor has many advantages such as high volume/sur-
residence time, lower volumetric energy content and low- face ratio, higher transport (e.g., heat and mass transfer) rates,
temperature operability [9,10]. Commonly, biodiesel is produced short diffusion distance, simplify process control and so on used
with batch reactors at a residence time that may take from one in the industrial process [21]. Sun et al. [22] used a microstructure
hour to several hours [11,12]. Due to the high residence time, many reactor with a 0.6 mm stainless steel capillary diameter for transe-
sterification of cotton seed oil with methanol. They evaluated
⇑ Corresponding author at: Chemical Engineering Department, Razi University, many parameters on transesterification and reached the best yield
Taghe Bostan, Kermanshah, Iran. Tel.: +98 8314274530; fax: +98 8314274542. of methyl ester at 120 °C with a residence time of 20 min. In an-
E-mail addresses: masoudrahimi@yahoo.com, m.rahimi@razi.ac.ir (M. Rahimi). other work by Sun et al. [23], it was shown that the obtained yield

0196-8904/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.enconman.2013.12.065
600 M. Rahimi et al. / Energy Conversion and Management 79 (2014) 599–605

of methyl ester in microstructured reactors could be over 94% at acid to neutralize extra base catalyst. The collected produced in
the residence time of less than 1 min. The experiments took place separation funnel was spited, into two phases, the FAME at the
with two different kinds of stainless steel capillary tube and a PTFE top and the glycerol at the bottom. Finally, the FAME phase was
tube that connected to a micromixer. In another study, transesteri- washed three times with distillated water at 80 °C for 5 min to re-
fication reaction of sunflower oil with ethanol in microreactors was move residuals catalyst and glycerol.
investigated by Richard et al. [24]. They proposed a kinetics model The FAME contents in samples were analyzed by gas chroma-
under various ethanol to oil molar ratios and validated it with their tography (Varian CP3800 Netherland), using a fused silica capillary
measured values. column (DB-WAX, 30 m_250 lm_0.25 lm, nominal) and a flame
Many methods were employed to design the experiments and ionization detector (FID). For this aim methyl laoranoate served
find out optimum condition in this process [25,26]. Among them, as the internal standard for GC. The analysis of FAME wt% for sam-
the response surface methodology is widely reported in literature. ples was done by method of Wang et al. [29] according to the fol-
Somnuk et al. [27] reported continuous acid-catalyzed esterifica- lowing equation:
tion of mixed crude palm oil using static mixer coupled with  
area of all FAME weight of reference
high-intensity ultrasonic irradiation. They used response surface FAME % ¼ 
methodology to optimize the two important reaction variables,
area of all reference weight of biodiesel sample
methanol and sulfuric acid concentrations. Jaliliannosrati et al.  100
[28] optimized biodiesel production from Jatropha curcas L. seed
using response surface methodology. Response surface methodol-
ogy (RSM) based on central composite design (CCD) was employed 2.3. Design of experiments and optimization
to analyze the influence of process variables on conversion of tri-
glycerides in a batch process. Biodiesel production depends on various parameters. Therefore,
In the present study, a microreactor with T-shaped junction was in order to decrease the number of tests, it is necessary to employ
used, as continuous flow reactor, for transesterification of soybean design of the experiment (DOE) to reduce costs and economize the
oil with methanol in presence of KOH. The influences the some time. Response surface methodology and central composite design
process variables such as; KOH concentration, molar ratio of meth- techniques have been reported in some research in literature for
anol to soybean oil, residence time and reaction temperature on design and optimization of many processes [30–33].
transesterification of triglycerides was investigated. In addition, In the present study, the Box–Behnken design was used to
optimization study was carried using response surface methodol- understand the effect of each variable on the continuous transeste-
ogy. The optimal transesterification conditions for producing the rification of soybean oil. Furthermore, using this technique it was
maximum methyl ester conversion and the relation between the tried to find optimum conditions and achieve the maximum bio-
parameters were determined using Box–Behnken experimental diesel yield.
design method. The factors selected for this optimization were KOH concentra-
tion, reactor temperature and molar ratio methanol to soybean oil.
The range and levels of them are listed in Table 1 and the maxi-
2. Materials and methods
mum values of the yield were taken as the responses in the exper-
iment design.
2.1. Materials
Analysis of variance (ANOVA) is a reliable method to analyze
and define the degree of certainty of experimental data [34]. Fur-
Soybean oil was purchased from a local market and employed
ther statistical analysis of the model was performed to evaluate
as feed for transesterification reactions. The additional materials
the ANOVA.
used during synthesis were: methanol (99.5%, Merck), potassium
The response variable (FAME %) was fitted with a full quadratic
hydroxide (85% pellets, Merck), hydrochloric acid (37%, Merck),
model in order to relate the FAME % to these variables. The form of
normal hexane (GC grade, >99%, Merck). Methyl laurate (Methyl
the mathematical model is shown in the Eq. (1):
dodecanoate, 99.7%) was obtained from (Sigma) as standard for
GC analysis. The Saponification index was (mg KOH/g oil) 191.88 X
3 X
3 X
2 X
3

and the mean molecular weight were calculated about 863.47 g/ Y yield ¼ b0 þ bi X i þ bii X 2i þ bij X ij ð1Þ
i¼1 i¼1 i¼1 j¼iþ1
mol.
where Y yield is the FAME %, Xi and Xij are the uncoded independent
2.2. Methods and experimental procedure variables, b0 is the offset term and bi, bii, bij are regression
coefficients.
Fig. 1 shows a schematic view of the microreactor rig. The mic-
roreactor included of a T-shaped plexiglass micromixer and a 3. Results and discussion
stainless steel microtube with an internal diameter of 0.8 mm. At
first step, in order to prepare methanol solution, it is necessary As discussed above, the operation conditions for the continuous
the calculated amounts of catalyst and methanol were mixed to- production of biodiesel by microreactor are quite wide. Therefore,
gether. The methanol/potassium hydroxide solution and oil were optimization of operational condition will be relatively important.
fed into two input channels of the studied T-micromixer by using In the following section, by using the Box–Behnken design of
a peristaltic pump (Qis™ DSP100). Both feeds mixed in the experiment, the effects of variables on the FAME % in the studied
micromixer and passed through the microtube. The resistance microreactor was explored. The corresponding Box–Behnken de-
time, the flow rates of the two feed streams and methanol solu- sign, experimental data for FAME % at the design points are sum-
tion/oil molar ratio were adjusted according to plan for experimen- marized in Table 2.
tal design. The effects of various operational parameters including; cata-
The microreactor was immersed in a water bath to adjust the lyst concentration, temperature and methanol/oil molar ratio on
reaction temperature. The product was collected in a tank and FAME % were studied. A residence time of 26 s was used, just as
for finishing the transesterification reaction, it was immersed in an example, for this analysis. In this reaction time, it was possible
an ice-water bath. Consequently, it was reacted with hydrochloric to see the effects of various operational parameters more clearly.
M. Rahimi et al. / Energy Conversion and Management 79 (2014) 599–605 601

Fig. 1. Schematic view of microreactor system.

of methyl esters conversion. In this figure, each point shows the


Table 1 mean percentage of methyl esters for all runs at the same molar ra-
Experimental range and levels of variable. tio of methanol to oil. As mentioned above, this analysis was car-
Variables Levels ried out at residence time of 26. It can obviously see that the
percentage of methyl esters increased almost linearly as the molar
1 0 1
ratio of methanol increased from 6:1 to 9:1. However, additional
Temperature (°C) 55 60 65
increase in the molar ratio of methanol from 9:1 to 12:1 caused
Catalyst concentration (wt.%) 0.6 1.2 1.8
Molar ratio (methanol/oil) 6:1 9:1 12:1 a slight decrease in the percentage of methyl esters. This behavior
might indicate that due to separation of glycerin and methyl ester,
an extra molar ratio of methanol can decrease the percentage of
3.1. Effect of molar ratio of methanol to oil
produced methyl esters.

Table 2
Experimental design and results. 3.2. Effect of reaction temperature
Experiment Temperature Catalyst Molar ratio FAME
no. (°C) concentration (methanol/oil) (%) Fig. 3 shows the main effect of temperature in percentage of
(wt.%) methyl esters conversion. The experiment was conducted at vari-
1 65 1.2 12 86.12
ety of temperatures at the fixed residence time about 26 s. In this
2 60 0.6 6 85.82 figure, each point shows the mean percentage of methyl esters
3 60 1.8 6 82.86 for all runs at the same temperature. It should be pointed out that
4 55 1.2 6 85.45 at low temperature, the reaction rate was slow and with increasing
5 55 1.2 12 89.77
temperature, the percentage of methyl esters remarkably in-
6 60 1.2 9 94.78
7 60 0.6 12 86.03 creased. It was found that the maximum percentage of methyl es-
8 55 0.6 9 88.64 ters for soybean oil using KOH was obtained at 60 °C. However, as
9 60 1.8 12 85.37 far as the temperature of reaction is higher than the boiling point
10 65 1.8 9 84.01 of methanol in laboratory pressure condition, further increase in
11 55 1.8 9 84.91
12 65 0.6 9 87.34
temperature to 65 °C resulted in the linearly decrease in percent-
13 65 1.2 6 79.24 age of methyl esters to 84%. Besides, due to high content of FFA
14 60 1.2 9 92.74 in oil, increase in reaction temperature accelerates the glycerides
15 60 1.2 9 93.09 saponification by the alkaline catalyst.

Based on transesterification reaction of oil, the stoichiometric 3.3. Effect of catalyst concentration
molar ratio of methanol to oil is 3 to 1. However, transesterification
reaction cannot be completed in this ratio and higher molar ratio In general, the catalyst concentration is vital in transesterifica-
should be used to force the reaction to be completed. Fig. 2 illus- tion process and cost of production depends on it and extra cata-
trates the effect of molar ratio of methanol to oil on percentage lyst will increase the complexity of product separation.
602 M. Rahimi et al. / Energy Conversion and Management 79 (2014) 599–605

KOH increased from 0.6 to 1.2 wt.% Moreover, increasing the


amount of KOH to 1.8 wt.% led to decrease in percentage of methyl
esters to 84%.
As discussed above, regarding to high content of FFA, these phe-
nomena cause more saponification of soybean oil at higher concen-
tration of KOH. The high percentage of 89% was obtained using
1.2 wt.% KOH.

3.4. Effect of the flow rate and the residence time under optimal
conditions

Due to short diffusion distance and high transport rate, the


reaction time in microreactor is very short and this is one of oper-
ational advantages of these deceives. Fig. 5 illustrates the effect of
various flow rates (the sum of flows oil and methanol), which
means different reactant residence times, on the percentage of
Fig. 2. Effect of molar ratio (methanol/oil) on FAME %. methyl esters conversion. The experiment was conducted at tem-
perature of 60 °C, KOH concentration of 1.2 wt.% and the molar ra-
tio (methanol/oil) of 9. The results indicate that percentage of
methyl esters conversion increased from 82.7% to 97.6% when
the residence time was increased from 20 to 180 s. The percentage
of methyl esters conversion first increased extremely with increase
in residence time from 20 to 60 s and after that increased slightly.
The maximum value of the methyl ester achieved almost 180 s and
transesterification of oil is almost complete. It can be seen that, at
residence time longer than 60 s, the percentage of methyl esters
conversion did not change significantly. Therefore, at the optimum
condition with minimum residence time a high percentage of
methyl esters were attained with the lowest biodiesel production
cost.
In order to further examine the effect of flow rate on conversion,
mixing inside of the T-shaped microchannel, a real picture cap-
tured from mixing in T-junction using a digital microscope,
(Dino-Lite, Taiwan.) In order to clearly show the boundary between
the two immiscible liquids, a small amount of dye was added to
the methanol. Fig. 6 shows the captured picture for low and high
Fig. 3. Effect of temperature (°C) on FAME %. flow rates. At low flow rate, although the residence time of reac-
tants in the reactor is sufficient but a straight borderline between
Therefore, the effect of different catalyst concentration at the fixed the two immiscible fluids is quite obvious. This indicates a non-
residence time about 26 s was investigated. Fig. 4 shows the effect efficient mixing in this flow rate. However, at high flow rate it
of catalyst concentration on percentage of methyl esters conver- can see that increase in flow rate cause a breaking in the borderline
sion. In this figure each point shows the mean percentage of and colored-methanol swap to the other side of the microchannel.
methyl esters for all runs at the same catalyst concentration. It was found that despite of low residence time, because of efficient
As shown in this figure, it can obviously seen that the percent- mixing in microreactor, the conversion percentage of methyl ester
age of methyl esters increased from 87% to 89% as the amount of was almost high.

Fig. 4. Effect of catalyst concentration (wt.%) on FAME %. Fig. 5. Effect of residence time (sec) on FAME %.
M. Rahimi et al. / Energy Conversion and Management 79 (2014) 599–605 603

Fig. 7. Pressure drop at different flow rate in microreactor.

The results show that the percentage of methyl esters conversion


approximately decreased linearly with increase in energy dissipa-
tion rate. As far as it is clear that higher dissipation rate related
to more efficient mixing, the results show that the residence time,
a time for reaction to take place, was more important than mixing
time. This is quite logical in slow reaction such as this transesteri-
fication reaction.
At the end, according to analysis of variable (ANOVA) of exper-
imental results a polynomial regression equation was developed
using the Box–Behnken design. The model expressed by Eq. (4)
and represents FAME % as a function of Molar ratio of methanol
to oil (x1), Catalyst concentration (x2) and reaction temperature
(x3).

Fig. 6. The fluid flow pattern inside T-shaped micromixer with two different flow FAME% ¼ 93:5367 þ 1:74X 1  1:335X 2  1:5075X 3
rates (a = 1 ml min1, b = 11 ml min1).
 4:7983X 21  3:7183X 22  3:5933X 23 þ 0:575X 1 X 2
þ 0:64X 1 X 3 þ 0:1X 2 X 3 : ð4Þ
3.5. Pressure drop and energy consumption
From statistical analysis, the values of 91.58% and 76.43% were
The pressure drop measurement in microchannel is vital in con- found for the determination coefficient (R2) and the adjusted coef-
tinuous processes. In order to evaluate the power consumption ficient (R2 adj.), respectively. This means that predicted values of
during the biodiesel production, the pressure drops across the FAME % from the model are in a good agreement with the experi-
channel (include micromixer and microtube) were measured at mental results and the model was reliable for biodiesel production
various feed flow rates. in this work. Furthermore, the significance of each variable in Eq.
The overall pressure drop along its length can be considered as (4) was calculated by p-values, which were listed in Table 3. The
follows: smaller magnitude of the p-value indicates the more significant
variable during experiment.
DPtotal ¼ DPmicromixer þ DPmicrotube : ð2Þ
Moreover, the pressure drop can be related to the energy dissi-
pation rate (e) for different conditions by following equation [35]:

DPQ
e¼ ð3Þ
qV
where, e is the energy dissipation rate per unit mass of fluid
(W kg1), DP is the pressure drop (Pa), Q the volumetric flow rate
(m3 s1) and V is the volume of fluid in the channel (m3). Therefore,
according to overall pressure drop across micromixer and micro-
tube the total the energy dissipation can be calculated.
Fig. 7 indicates an increase in pressure drop with increase in the
reactant flow rate from 1 to 11 ml min1. The experimental results
indicate that with increasing flow rate the pressure drop increases
linearly. Therefore, it can be concluded the flow regime was lami-
nar at the range of employed flow rates.
In Fig. 8, the obtained FAME % versus the energy dissipation rate
is shown. The energy dissipation rate was calculated based on flow
rate energy and pressure drop across the inlet and the outlet sides. Fig. 8. Comparison FAME % as a function of the energy dissipation ratio.
604 M. Rahimi et al. / Energy Conversion and Management 79 (2014) 599–605

Table 3
The values calculated by ANOVA.

Variable DF Seq SS Adj SS Adj MS F P


Molar ratio(methanol/oil) 1 24.221 24.221 24.2208 6.00 0.058
Catalyst concentration (wt.%) 1 14.258 14.258 14.2578 3.53 0.119
Temperature (°C) 1 18.180 18.180 18.1804 4.50 0.087
Molar ratio(methanol/oil)  molar ratio(methanol/oil) 1 68.263 85.012 85.0117 21.06 0.006
Catalyst concentration (wt.%)  catalyst concentration (wt.%) 1 44.003 51.050 51.0499 12.65 0.016
Temperature (°C)  temperature (°C) 1 47.675 47.675 47.6752 11.81 0.019
Molar ratio (methanol/oil)  catalyst concentration (wt.%) 1 1.322 1.322 1.3225 0.33 0.592
Molar ratio(methanol/oil)  temperature (°C) 1 1.638 1.638 1.6384 0.41 0.552
Catalyst concentration (wt.%)  temperature (°C) 1 0.040 0.040 0.0400 0.01 0.925
Residual error 5 20.183 20.183 4.0365
Lack-of-fit 3 17.802 17.802 5.9342 4.99 0.172
Pure error 2 2.380 2.380 1.1900
Total 14 239.783

Table 4
Comparison between obtained FAME % in different kinds of reactors.

Reactor FAME (%) Residence time (min) T (°C) Oil Catalyst


Stirred-tank reactor [19] 58.8–97.3 40–70 60 Palm oil KOH
Microwave heating reactor [36] 94.4–95.25 0.56 50 Soybean oil KOH
Ultrasonic [37] 72–96 10–30 38 Palm oil KOH
Reactive distillation reactor [38] 41.5–97 2.67–6.67 55 Canola oil KOH
Present work 82–89 0.43 60 Soybean oil KOH

Based on the linear effect and considering the p-value (less than at methanol/oil molar ratio, 9:1, catalyst concentration, 1.2 wt%,
0.05), the molar ratio (methanol/oil) term had the most significant and the reaction temperature of 60 °C with a residence time of
effect on the percentage of methyl ester conversion and in the sec- 26 s. Furthermore, 98% conversion were obtained at optimum con-
ond place reaction temperature was important. All quadratic terms dition with longer residence time of 180 s. In order to compare the
in Eq. (4) are significant terms in increasing the FAME %. However, importance of mixing and reaction residence time, pressure drop
the molar ratio (methanol/oil) is the most important term among across the microreactor at various flow rates were measured. The
them. From Table 3, one can say that the interaction terms has results showed that the residence time is more important than
not any significant effect on the FAME %. Therefore, based on the mixing time, which is quite logical in slow transesterification
p-value, which indicate significant terms, some terms can be de- reaction.
leted and the correlation can be shortened with almost 2–3% differ-
ences in the predicted values as follows:

FAME% ¼ 93:5367  4:7983X 21  3:7183X 22  3:5933X 23 : ð5Þ References

[1] Demirbas MF, Balat M. Recent advances on the production and utilization
3.6. Comparison with conventional batch reactors trends of bio-fuels: a global perspective. Energy Energy Convers Manage
2006;47(15):2371–81.
[2] Bozbas K. Biodiesel as an alternative motor fuel: production and policies in the
In another part of this study, a comparative work has been European Union. Renew Sustain Energy Rev 2008;12:542–52.
undertaken to compare advantages of this continuous reactor sys- [3] Demirbas A, Demirbas I. Importance of rural bioenergy for developing
countries. Energy Convers Manage 2007;48:2386–98.
tem with conventional reported batch systems [19,36–38].
[4] Demirbas A. Comparison of transesterification methods for production of
Although continuous and batch rectors cannot be compared di- biodiesel from vegetable oils and fats. Energy Convers Manage
rectly but the order of magnitude of residence time and yields of 2008;49:125–30.
[5] Baroutian S, Aroua MK, Raman AAA, Sulaiman NMN. Density of palm oil-based
FAME % can allow a qualitative judgment. The results of this com-
methyl ester. J Chem Eng Data 2008;53:877–80.
parison, in terms of FAME %, residence time and the operational [6] Vicent G, Coteron A, Martinez M, Aracil J. Application of the factorial design of
temperature, are listed in Table 4. As can be seen from this table, experiments and response surface methodology to optimize biodiesel
the FAME % and the time of reaction in microreactors is quite lower production. Ind Crops Prod 1998;8:29–35.
[7] Soriano JN, Venditti R, Argyropoulos D. Biodiesel synthesis via homogeneous
compared with conventional rectors. This result reveals the advan- Lewis acid-catalyzed transesterification. Fuel 2009;88:560–5.
tage of using the proposed microreactor system over conventional [8] Maceiras R, Vega M, Costa C, Ramos P, Márquez MC. Enzyme deactivation
processes. during biodiesel production. Chem Eng J 2011;166(1):358–61.
[9] Ma F, Hanna MA. Biodiesel production: a review. Bioresour Technol
1999;70:1–15.
4. Conclusion [10] Canakci M, Erdil A, Arcaklioğlu E. Performance and exhaust emissions of a
biodiesel engine. Appl Energy 2006;83:594–605.
[11] Campanelli P, Banchero M, Manna L. Synthesis of biodiesel from edible, non-
In the present study, experimental work on continuous biodie- edible and waste cooking oils via supercritical methyl acetate
sel production using a T-shaped microreactor was carried out. The transesterification. Fuel 2010;89:3675–82.
[12] Patil PD, Deng S. Optimization of biodiesel production from edible and non-
main issue in this study is to find the effects of methanol/oil molar
edible vegetable oils. Fuel 2009;88:1302–6.
ratio, catalyst concentration, reaction time and reaction tempera- [13] Pengmei Lu, Yuan Zhenhong, Li Lianhua, Wang Zhongming, Luo Wen. Biodiesel
ture on the percentage of methyl esters conversion (FAME %). from different oil using fixed-bed and plug-flow reactors. Renew Energy
The optimal status for biodiesel production was defined with the 2010;35(1):283–7.
[14] Falahati H, Tremblay AY. The effect of flux and residence time in the
aid of the Box–Behnken method of experimental design. The bio- production of biodiesel from various feedstocks using a membrane reactor.
diesel with best percentage of methyl esters, 89%, was produced Fuel 2012;91(1):126–33.
M. Rahimi et al. / Energy Conversion and Management 79 (2014) 599–605 605

[15] Hanh HD, Dong NT, Starvarache C, Okitsu K, Maeda Y, Nishimura R. [28] Jaliliannosrati H, Amin NAS, Talebian-Kiakalaieh A, Noshadi I. Microwave
Methanolysis of triolein by low frequency ultrasonic irradiation. Energy assisted biodiesel production from Jatropha curcas L. seed by two-step in situ
Convers Manage 2008;49(2):276–80. process: Optimization using response surface methodology. Bioresour Technol
[16] Zhang H, Ding J, Zhao Z. Microwave assisted esterification of acidified oil from 2013;136:565–73.
waste cooking oil by CERP/PES catalytic membrane for biodiesel production. [29] Wang Y, Ou S, Liu P, Xue F, Tang S. Comparison of two different processes to
Bioresour Technol 2012;123:72–7. synthesize biodiesel by waste cooking oil. Mol Catal A: Chem
[17] Yin JZ, Xiao M, Song JB. Biodiesel from soybean oil in supercritical methanol 2006;252(1):107–12.
with co-solvent. Energy Convers Manage 2008;49(5):908–12. [30] Torrades F, Saiz S, García-Hortal JA. Using central composite experimental
[18] Noureddini H, Harkey D, Medikonduru V. Continuous process for the design to optimize the degradation of black liquor by Fenton reagent.
conversion of vegetable oils into methyl esters of fatty acids. J Am Oil Chem Desalination 2011;268(1):97–102.
Soc 1998;75(12):1775–83. [31] Tsapatsaris S, Kotzekidou P. Application of central composite design and
[19] Darnoko D, Cheryan M. Continuous production of palm methyl esters. J Am Oil response surface methodology to the fermentation of olive juice by
Chem Soc 2000;77(12):1269–72. Lactobacillus plantarum and Debaryomyces hansenii. Int J Food Microbiol
[20] Peterson C, Cook J, Thompson J, Taberski J. Continuous-flow biodiesel 2004;95:157–68.
production. Appl Eng Agric 2002;18(1):5–11. [32] Wang H, Liu Y, Wei S, Yan Z. Application of response surface methodology to
[21] Kumar V, Paraschivoiu M, Nigam KDP. Single-phase fluid flow and mixing in optimise supercritical carbon dioxide extraction of essential oil from Cyperus
microchannels. Chem Eng Sci 2011;66(7):1329–73. Rotundus Linn. Food Chem 2012;132:582–7.
[22] Sun P, Sun J, Yao J, Zhang L, Xu N. Continuous production of biodiesel from high [33] Chen Y, Zhao L, Liu B, Zuo S. Application of response surface methodology to
acid value oils in microstructured reactor by acid-catalyzed reactions. Chem optimize microwave-assisted extraction of polysaccharide from tremella. Phys
Eng J 2010;162(1):364–70. Procedia 2012;24:429–33.
[23] Sun P, Wang B, Yao J, Zhang L, Xu N. Fast synthesis of biodiesel at high [34] Montgomery DC. Design and analysis of experiments. New York: John Wiley
throughput in microstructured reactors. Ind Eng Chem Res 2009;49:1259–64. and Sons; 2001.
[24] Romain R, Sophie TR, Laurent P. Modelling the kinetics of transesterification [35] Falk L, Commenge JM. Performance comparison of micromixers. Chem Eng Sci
reaction of sunflower oil with ethanol in microreactors. Chem Eng Sci 2010;65:405–11.
2013;87:258–69. [36] Barnard TM, Nicholas EL, Matthew BB, Lauren MS, Benjamin AW. Continuous
[25] Al-Dawody MF, Bhatti SK. Optimization strategies to reduce the biodiesel NOx flow preparation of biodiesel using microwave heating. Energy Fuel
effect in diesel engine with experimental verification. Energy Convers Manage 2007;21:1777–81.
2013;68:96–104. [37] Stavarache C, Vinatoru M, Maeda Y, Bandow H. Ultrasonically driven
[26] Tashtoush GM, Al-Widyan MI, Al-Jarrah MM. Experimental study on continuous process for vegetable oil transesterification. Ultrason Sono Chem
evaluation and optimization of conversion of waste animal fat into biodiesel. 2007;14:413–7.
Energy Convers Manage 2004;45(17):2697–711. [38] He BB, Singh AP, Thompson JC. Experimental optimization of a continuous flow
[27] Somnuk K, Smithmaitrie P, Prateepchaikul G. Optimization of continuous acid- reactive distillation reactor for biodiesel production. Trans ASAE
catalyzed esterification for free fatty acids reduction in mixed crude palm oil 2005;48:2237–43.
using static mixer coupled with high-intensity ultrasonic irradiation. Energy
Convers Manage 2013;68:193–9.

You might also like