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

net/publication/308152114

Optimization of the first step transesterification of refined palm oil using

solid sodium methoxide catalyst

Conference Paper · October 2016

DOI: 10.13140/RG.2.2.17398.19521

CITATIONS READS

2 896

4 authors:

Dang Nguyen Thoai Chakrit Tongurai

Quy Nhon University Prince of Songkla University
11 PUBLICATIONS   24 CITATIONS    63 PUBLICATIONS   753 CITATIONS   

SEE PROFILE SEE PROFILE

Kulchanat Prasertsit Songtham Photaworn

Prince of Songkla University Prince of Songkla University
29 PUBLICATIONS   147 CITATIONS    7 PUBLICATIONS   20 CITATIONS   

SEE PROFILE SEE PROFILE

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

2 step transesterification via 1st homogeneous catalyst transesterification and 2nd heterogeneous catalyst View project

Microwave heating for small-scale milling process View project

All content following this page was uploaded by Dang Nguyen Thoai on 17 September 2016.

The user has requested enhancement of the downloaded file.

23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

Optimization of the first step transesterification of

refined palm oil using solid sodium methoxide
catalyst
Dang Nguyen Thoai*, Chakrit Tongurai, Kulchanat Prasertsit, Songtham Photaworn
Department of Chemical Engineering, Faculty of Engineering
Prince of Songkla University
Songkhla 90112, Thailand
*Corresponding author’s email: dangnguyenthoai@qnu.edu.vn

Abstract: The two-step transesterification process in biodiesel indicated its strong point, especially for feedstock with free
production has been studied in the previous studies. However, fatty acid (FFA) below 2 wt.%, reported to proceed about 4000
there has not had any appropriate method to optimize for the times of reaction speed faster than acid homogeneous catalyst
first step of transesterification by base homogeneous catalyst. In [3]. The common base homogeneous catalysts include sodium
this novel study, based on response surface methodology (RSM), hydroxide (NaOH), potassium hydroxide (KOH), sodium
a mathematical model was established which indicated the methoxide (CH3ONa) and potassium methoxide (CH 3OK) [5].
impact of independent experimental conditions as well as the
relationship between these variables (methanol/refined oil molar These popular catalysts can catalyze reaction at smooth
ratio, catalyst content, time, temperature) to the methyl esters condition (low temperature, atmospheric pressure) but high
content. This model was significant by comparison with both conversion can be obtained in a minimum time. In all of these
experimental value and predicted number of the methyl ester base homogeneous catalysts, CH3ONa has been demonstrated
content, and shown strong agreement for all four experimental to be the most effective catalyst [5-7]. Dias et al. [7] reported
parameters. The methyl ester content of 80% as the target of this that the high ester content (97%) could be reached by using
stage was obtained in optimal conditions: methanol/refined oil 0.6 wt.% of CH3ONa in a quite slight condition
molar ratio (5.30:1), catalyst content (0.29 wt.% of CH3ONa to (MeOH/refined oil molar ratio 6:1, reaction temperature 60oC,
refined palm oil), time (40 min), temperature (57 oC). The results reaction time 60 min).
also indicated that catalyst content is the most significant factor
for the first step transesterification. However, the most remarkable disadvantage of using
base homogeneous catalyst is very easy to occur the
Keywords: Biodiesel, two-step transesterification, first step saponification reaction, especially for high FFA feedstock
transesterification, sodium methoxide, RSM. [5,8]. The saponification reaction is a significant side-reaction
in the transesterification process. This reaction is occurred
I. INTRODUCTION
between FFA and base homogeneous catalyst and consumes
The hotter and hotter development of the world economy the amount of catalyst to produce soaps. That means in order
requires the higher and higher requirement about the fuel. to ensure the stability of transesterification process, a higher
However, petroleum-based fuel increases greenhouse gas amount of catalyst should be added, but this also accretes the
emissions and leads to the environmental hazards and the saponification reaction. Soap formation inhibits the separation
global warming, the studying and using of nonpetroleum- of biodiesel, glycerin and washed water. The soap formation is
based fuel as alternative resources are more appreciated [1]. also the main factor of yield loss.
Biodiesel, an alternative renewable fuel, is a mixture of mono
alkyl esters of fatty acids which is gained via the Moreover, one of the drawbacks of base homogeneous
transesterification reaction of triglycerides (TG) – main catalyst is the existence of the residual amount of catalyst. If
component in many feedstocks (vegetable oils and animal the transesterification process is carried out at conventional
fats) – with methanol, ethanol or other suitable alcohols in the conditions (atmospheric pressure, 65 oC of temperature) with
presence of a relevant catalyst [1,2]. In comparison with the excess of methanol, then it is so difficult to guarantee that
petroleum diesel, biodiesel has many outstanding advantages, alkaline homogeneous catalyst catalyzed the reaction
such as renewability, biodegradability, non-toxic for completely. Residual catalyst leads many challenges for
environment, high safety as well as acceptable properties for catalyst removal technology. Therefore, it increases the
diesel engine [1–4]. production cost and the final product, respectively [9].

The catalyst plays an important role in transesterification In order to overcome the demerits of base homogeneous
reaction. In both of the recently studies and industry catalyst, solid acid catalyst was studied and used in
production, the use of base homogenous catalysts has experimental scale. One of its merits is able to be insensitive
23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

to FFA and water content in the feedstock [5]. This advantage stirring used magnetic bar and stirring speed of 600 rpm,
makes acid heterogeneous catalyst to be suitable for various worked at atmospheric pressure, refluxed by water at 20 oC to
feedstock oils (edible oil, non-edible oil, waste oil) and to be condense the methanol vapor. RPO with FFA content about
easy to separate residue catalyst due to no soap formation. 0.1 wt.% (to RPO) was used as a raw feedstock. The feedstock
However, low conversion is the most disadvantage of using oil was preheated until the temperature reached to the
this catalyst. Ji et al. [10] shown that in normal experimental requested value. And then, the mixture of methanol and
condition (MeOH/Oil molar ratio 8:1, 3 wt.% Amberlyst 15 to catalyst was added. The beginning time for the reaction was
oil, reaction temperature 60 oC, reaction time 90 min), the counted at the moment all of methanol and catalyst have been
conversion was only 25%. In a research later of Nezahat et al. entered to the reactor. After finished reaction, the reaction
[11], by increasing some experimental conditions ((MeOH/Oil product was transferred to separatory funnel and settled within
molar ratio 12:1, 3 wt.% Amberlyst 15 to oil, reaction 60 min to separate into two phases, methyl ester phase and
temperature 65 oC, reaction time 9 h), ester content was about glycerol phase. When glycerol phase was removed from the
78%. This value of ester content did not respond to the quality methyl ester phase, the methyl ester phase was washed with
of commercial biodiesel seen in standards such as ASTM D- hot water, at 70 oC, about three times without stirring and three
6571 or EN 14214. times with stirring. The washed methyl esters were dried by
the heating in 90 min, at 80 oC. The last biodiesel was weighed
The technology of doing two-step reaction in biodiesel to determine the methyl esters content.
production process was shown in some previous studies
[8,12,13]. This process can be done by homogeneous catalyst All the experimental runs were done three times to
as well as heterogeneous catalyst. The most significant estimate its errors. Experiments were designed at various
advantage of this technology is to decrease experimental conditions such as, methanol/RPO molar ratio (4.85.8),
conditions and to gain the efficiency of total transesterification catalyst content (0.250.50 wt.% of CH3ONa to RPO),
process. According to the study of Noureddini et al. [14] about reaction time (2060 min) and reaction temperature (5065
the kinetics of transesterification of soybean oil, o
C).
approximately 80% conversion was obtained after
transesterification process in conventional experimental C. Experimental designs
conditions. Therefore, 80% of ester content is appropriate Response surface methodology (RSM) is one of
target for the first step transesterification by base significant methods which more details regression models are
homogeneous catalyst. used to determine response behavior. This statistical method
was utilized for analyzing of the experimental data using EED
Two-step transesterification via the first step
software in MS Excel [15]. After loading EED, an additional
transesterification using homogeneous base catalyst and the
menu option, DOE (Design of Experiment), will become
second transesterification using heterogeneous catalyst is our
available in the Main Menu of MS Excel (Menu Add-Ins). The
interest. In this work, by using RSM as a statistical method,
Central Composite Design (CCD) was used to find the
the novel point is to analyze the impact of experimental
optimum conditions for the requested methyl esters content.
conditions to the methyl ester content and to determine the
The CCD incorporates five levels (coded –α, –1, 0, +1, +α) in
optimal condition for the first step transesterification process.
which axial points (±α) for a factor and 0 for all other factors.
II. MATERIALS AND METHODS In addition, center points were coded as 0 and used to estimate
pure error. The most important factors for transesterification
A. Materials including molar ratio (X1), catalyst content (X2), reaction
Refined palm oil (RPO) in this project was purchased time (X3) and reaction temperature (X4) were chosen as
from Morakot Industry Public Co. Ltd. (Thailand). The base independent variables and the methyl esters content was the
homogeneous catalyst, solid sodium methoxide (CH3ONa, 96 dependent variable. The experimental limit and coded levels
wt.%), was supplied by Dezhou Long Teng Chemical Co. Ltd. of independent factors for this study were shown in Table 1. A
(China) whereas methanol (CH3OH, 99.8 wt.%) and H2SO4 list of 29 experiments including the 24 factorial runs, 8 runs
(98 wt.%) were obtained from Labscan Asia Co. Ltd. for axial points and 5 runs for center points were carried out.
(Thailand). Other important chemicals, sodium periodate A smooth interpolating function is usually used, and it
(NaIO4) was acquired from Fisher Chemical (UK), NaOH was generally includes quadratic and interaction effects. The
obtained from Merck (Germany) and bromothymol blue were general form of second order polynomial regression model
provided by Ajax Finechem (Australia). equation is expressed as follows:
B. Transesterification reaction procedure and phase
∑ ∑ ∑ ∑
separation
Transesterification reaction of RPO with methanol in the where Y is the predicted response (methyl esters content); β0,
presence of CH3ONa as a base homogeneous catalyst was βi, βii, βij are the regression coefficients (β0 is referred to as the
carried out in a 0.5 L three-necked flask, with magnetic
23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

constant term, βi is a linear term, βii is a quadratic term and βij date to (1). Based on the coded factors and by removing the
is an interaction term); Xi, Xj are coded independent factors. insignificant model terms, the adjusted regression model was
shown in (2):
TABLE 1. LIMIT ANF CODED LEVELS OF INDEPENDENT IN
TRANSESTERIFICATION REACTION Esters content (%) = 19.90 + 184.12X2 – 225.76X22 +
0.07437X1X3 – 0.641X2X3 + 1.019X2X4 (2)
Factor Limit and coded level
Independent Symbol Dimension –α –1 0 +1 +α The analysis of variance (ANOVA) was summarized in
variable
Table 3. The fit of the designed model with the experimental
results was tested due to F-value, P-value, R2 as well as lack
Molar ratio X1 mol/mol 4.80 5.05 5.30 5.55 5.80
of fit (LOF). As shown in Table 3, the F-value (45.40) of
Catalyst content X2 wt.% 0.25 0.31 0.38 0.44 0.50 suggested model was much higher than Fcrit (F(0.05,5,23) =
Time X3 min 20 30 40 50 60 9.87) and the P-value was very low (3.63E-11). These values
o
Temperature X4 C 50 54 58 61 65 indicated clearly that the corresponding model was significant
After the completion of the design of experiments and at the 95% confidence level. The performance of the model
doing 29 experimental runs, ER software is used for the was also based on using the coefficient of multiple
essential regression, including Multiple Regression and determination or correlation coefficient, R2, which was
Polynomial Regression. In order to use this software, ER is determined to be 0.908 at 95% of confidence interval.
loaded into MS Excel. After that, a new additional menu Correlation coefficient, R2, illustrated that 90.8% of the
option, Regress, becomes available in the Main Menu of MS methyl esters content changing was affected by independent
Excel. variables including methanol to RPO molar ratio, catalyst
content, reaction time and reaction temperature. About 9% of
D. Approximate analysis of total glyceride and ester content this changing was caused by unidentified factors (random
The ester content determination was carried out error). In addition, the LOF (0.06436) was higher than 0.05
according to the Thailand Patty Patent No. 5060 [16] as well indicated that the LOF was insignificant relative to the pure
as from Pissarello et al. [17]. Tongurai [16] indicates that total error [18]. All things considered this model was quite
glyceride content in biodiesel can be determined from appropriate to predict the methyl esters content.
transesterification using microwave radiation technique. As has been shown in Table 3, each term in the adjusted
Residues glycerides in a biodiesel will react with methanol in regression model was also checked in order to evaluate how
the presence of potassium methoxide to produce methyl esters well the significance and its interaction to the methyl esters
and glycerol. The amount of glycerol can be defined by using content. A P-value less than 0.05 implies that the relevant
a correlation curve. Then the total glyceride content can be model term is significant. From Table 3, it was noted that after
converted into ester content by subtracting the value from 100 removing insignificant model terms, with approaching the
wt.%. In the other proximate method [17], an exact amount of linear, quadratic and combined impacts, linear term of catalyst
biodiesel sample will be weighed and the glyceride in content (X2), quadratic term of catalyst content (X2 2),
biodiesel will react with sodium methoxide solution (in interaction term between molar ratio and time (X1X3) and
methanol) in a batch reactor. After finished reaction, the interaction terms of catalyst content with time (X2X3) as well
amount of glycerin, stand-in of glyceride content, is as with temperature (X2X4) were significant. Almost these
determined by titrimetric method. Approximate ester content data proved the catalyst content is the most important factor in
transesterification process by base homogeneous catalyst. Due
is calculated by subtracting the glyceride content from 100
to the described coefficients in Eq. (2), catalyst content has an
wt.%. active effect on the methyl esters content. That means the
III. RESULTS AND DISCUSSIONS increasing of catalyst content accelerates the speed of the
transformation from triglycerides to methyl esters. This
A. Regression model and statistical analysis for methyl esters conclusion can be seen in some previous studies [19-21]. In
content the considering for other interaction terms, the combination of
It is clear that the methyl ester content obtained from catalyst content and temperature (X2X4) was the most
transesterification reaction is affected by four independent significant with the lowest P-value (9.8E-06).
variables. In this study, the mathematical relationships Fig. 1a described the relationship between the predicted
between the methyl esters content as a response and four values and the experimental values of the methyl esters
reaction conditions, methanol to RPO molar ratio, catalyst content. This relationship was nearly linear, that means to be
content, time and temperature were estimated. Twenty nine sure about the high significance of the model. One more
experiments were established and the results were analyzed by important to realize the role both R2 and adjusted R2 in the
using RSM for the methyl esters content as shown in Table 2. evaluation the significance of the suggested model, large
By applying multiple regression analysis in Table 2, the full differences between R2 and adjusted R2 demonstrate the
factorial central composite design was used to fit the obtained insignificance of the model [22]. From the data in Table 3, this
23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

deviation was only 2.25%. Therefore, the suggested model cases has considerable residuals [23]. Almost the standard
was right description of the process. residuals should be limited in the interval of ±3.00. As can be
seen in Fig. 1b, there were not any value outside the
The outlier t plot for all experimental runs was shown in prescribed interval (±3.00). It means that the suggested model
Fig. 1b. The outlier t plot specifies the amplitude of the is relevant with all the data.
residuals for each case to determine if any of experimental
TABLE 2. THE CODED INDEPENDENT FACTORS, EXPERIMENTAL
RESULTS AND PREDICTED VALUES (a)
Run Independent variables Ester content (%)
Residuals
X1 X3 X4 Experimental Predicted
X2 (wt.%)
(mol/mol) (min) (oC) value value
1 –1 (5.05) –1 (0.31) –1 (30) –1 (54) 78.35 78.01 0.34
2 +1 (5.55) –1 (0.31) –1 (30) –1 (54) 76.66 77.52 –0.86

3 –1 (5. 05) +1 (0.44) –1 (30) –1 (54) 85.15 85.07 0.08

4 +1 (5. 55) +1 (0.44) –1 (30) –1 (54) 83.88 85.05 –1.17

5 –1 (5. 05) –1 (0.31) +1 (50) –1 (54) 81.21 80.80 0.41

6 +1 (5. 55) –1 (0.31) +1 (50) –1 (54) 82.57 81.92 0.65

7 –1 (5. 05) +1 (0.44) +1 (50) –1 (54) 84.70 86.39 –1.69

8 +1 (5. 55) +1 (0.44) +1 (50) –1 (54) 87.70 87.99 –0.29

9 –1 (5. 05) –1 (0.31) –1 (30) +1 (61) 80.20 80.49 –0.29

10 +1 (5. 55) –1 (0.31) –1 (30) +1 (61) 82.75 81.80 0.95

(b)
11 –1 (5. 05) +1 (0.44) –1 (30) +1 (61) 84.95 86.35 –1.40

12 +1 (5. 55) +1 (0.44) –1 (30) +1 (61) 87.15 88.14 –0.99

13 –1 (5. 05) –1 (0.31) +1 (50) +1 (61) 84.11 83.68 0.43

14 +1 (5. 55) –1 (0.31) +1 (50) +1 (61) 85.96 86.61 –0.65

15 –1 (5. 05) +1 (0.44) +1 (50) +1 (61) 88.37 88.08 0.29

16 +1 (5. 55) +1 (0.44) +1 (50) +1 (61) 90.38 91.48 –1.10

17 – (4.80) 0 (0.38) 0 (40) 0 (58) 84.75 84.49 0.26

18 + (5.80) 0 (0.38) 0 (40) 0 (58) 88.48 87.41 1.07

19 0 (5.30) – (0.25) 0 (40) 0 (58) 74.85 76.00 –1.15

20 0 (5.30) + (0.5) 0 (40) 0 (58) 90.41 87.93 2.48 Fig.1 (a) Correlation between experimental and predicted
21 0 (5.30) 0 (0.38) – (20) 0 (58) 82.76 81.75 1.01
methyl esters content. (b) The outlier t plot for all cases.

22 0 (5.30) 0 (0.38) + (60) 0 (58) 88.19 87.88 0.31

23 0 (5.30) 0 (0.38) 0 (40) – (50) 82.24 81.63 0.61

24 0 (5.30) 0 (0.38) 0 (40) + (65) 88.32 87.60 0.72

25 0 (5.30) 0 (0.38) 0 (40) 0 (58) 85.07 85.75 –0.68

26 0 (5.30) 0 (0.38) 0 (40) 0 (58) 85.70 85.75 –0.05

27 0 (5.30) 0 (0.38) 0 (40) 0 (58) 85.26 85.75 –0.49

28 0 (5.30) 0 (0.38) 0 (40) 0 (58) 86.42 85.75 0.67

29 0 (5.30) 0 (0.38) 0 (40) 0 (58) 86.31 85.75 0.56

23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

TABLE 3. ANOVA RESULTS FOR THE ADJUSTED REGRESSION MODEL

Sum of squares Degree of Mean F-Signif/ P

Source/ Term Coefficient F value Remarks
(SS) freedom (DF) squares (MS) value
Regression 358.65 5 71.73 45.40 3.63E-11 Significant
β0 19.90 0.04821 Significant
β2 184.12 1 0.000896 Significant
β22 –225.76 1 0.00122 Significant
β13 0.07437 1 0.00147 Significant
β23 –0.641 1 0.03945 Significant
β24 1.019 1 9.8E-06 Significant
Residual 36.34 23 1.580 5.0178
Lack of fit error 34.88 19 1.836 0.06436 Not significant
Pure error 1.46 4 0.366
Total 394.99 28
Correlation coefficient: R2 = 0.908 (R2 adjusted = 0.888)
Fcrit = F(0.05,5,23) = 9.87

B. Interaction effects of the factors on methyl esters content (a)

1. Interaction effect of catalyst content and temperature on

methyl esters content
The effects of catalyst content and temperature on methyl
esters content were investigated with keeping the MeOH/RPO
molar ratio and reaction time at the medium values, 5.30
mol/mol and 40 min, respectively. This influence was shown
by the response surface and contour plots in Fig. 2. The slope
of the contours will decide the degree of the interaction of
process factors to methyl esters content. The higher slope of
the contour of catalyst content indicated great influence of this
variable in comparison with reaction temperature. It was
realized that catalyst loading significantly impacted on methyl
esters content, especially at the concentration higher than 0.32
wt.%. By using at least 0.32 wt.% of CH3ONa (to RPO),
methyl esters content clearly increased without depending on (b)
the experimental temperature limitation (50 – 65 oC). In other
words, methyl esters content was enhanced by enlargement
catalyst concentration at any temperature range. The same
conclusion can be seen in some previous results [19,24].
However, at lower than catalyst loadings (less than 0.32
wt.%), the effect of reaction temperature on methyl esters
content was appreciable. The increasing rate of methyl esters
content followed the growth limit of reaction temperature,
although this effect was slow because of lower slope of the
contour.

Fig. 2. Response surface and contour plots for interaction effect of catalyst
content and reaction temperature (MeOH/RPO molar ratio: 5.3 and reaction
time: 40 min).
23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

(a) (b)

Fig. 3. Response surface and contour plots for interaction effect of molar ratio and reaction time (catalyst content: 0.38 wt.% and reaction temperature: 58 oC).

2. Interaction effect of molar ratio and reaction time on methyl performed by the response optimization built in ER software
esters content loaded into MS Excel. The optimum conditions to gain 80% of
methyl esters content were indicated at the catalyst content of
In Fig. 3, response surface and contour plots for the 0.29 wt.%, MeOH/RPO molar ratio of 5.30:1, reaction time of
interaction variable of molar ratio and reaction time were 40 min and reaction temperature of 57 oC.
clearly shown with the constant value of catalyst content and
reaction temperature, 0.38 wt.% and 58 oC, respectively. By IV. CONCLUSIONS
comparison the difference about the slope of contour, it was
This work has been successful in constituting the
indicated that the reaction time had higher influence than the
feedstock with 80% of methyl esters for next steps in the
molar ratio. The methyl esters content increased significantly
process of manufacturing commercial biodiesel. This
with increasing reaction time at the same MeOH/RPO molar
feedstock is very convenient for second step transesterification
ratio due to higher slope of the contour. In contrast, the effect
process by heterogeneous catalyst.
of MeOH/RPO molar ratio on methyl esters content was not as
much as the reaction time. This can be explained by the good Our study reaffirms the important role of the response
solubility of MeOH in the presence of methyl esters and surface methodology (RSM) in optimizing the experimental
intermediate compounds (mono_ and di_glycerides). Thus the conditions in order to gain the expected content of methyl
importance of MeOH solubility ability in RPO is diminished. esters. By the considering the objective of the first step
The similar explanation was seen in some previous results transesterification process, the optimal conditions were
[25,26]. indicated, including catalyst content (0.29 wt.%), MeOH/RPO
molar ratio (5.30:1), reaction time (40 min) and reaction
However, with considering the objective of this first
temperature (57 oC).
stage (80% of esters content), the role of this interaction effect
of molar ratio and reaction time on methyl esters content was This novel research can be applied for different feedstock
not as significant as the interaction effect of catalyst content oils, such as edible oil, non-edible oil as well as waste cooking
and reaction temperature. As shown in Fig. 3, the methyl oil. Further study is recommended to determine the optimal
esters content was absolutely higher than the target of this conditions for the second step transesterification process by
stage. One more explanation, from Table 3, interaction term of heterogeneous catalyst for the higher biodiesel yield and the
catalyst content with temperature (X2X4) was much higher lower price of commercial biodiesel.
than interaction term of molar ratio with time (X1X3), 9.8E-06
and 0.00147, respectively. ACKNOWLEDGMENT
C. Optimization of process factors for the first step The authors would like to extend our gratitude to
transesterification process Graduate School of Prince of Songkla University for
supporting the fund under Thailand’s Education Hub for
Optimization is defined as the process of finding Southern Region of ASEAN countries (TEH-AC). Our thanks
optimized settings of the regressions in the model in order to are also extended to Department of Chemical Engineering and
obtain a predefined output or response value [16]. In order to Specialized Research and Development Center for Alternative
determine the optimum conditions for first stage Energy from Palm Oil and Oil Crops, Faculty of Engineering,
transesterification process, numerical optimization was Prince of Songkla University.
 23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

REFERENCES Wang, J.W. Zhang, and D.P. Yang, “A two-step

[1] S. Soltani, U. Rashid, R. Yunus, and Y.H. Tayfig- biodiesel production process from waste cooking oil
Yap, “Synthesis of biodiesel through catalytic via recycling crude glycerol esterification catalyzed
transesterification of various feedstocks using fast by alkali catalyst,” Fuel Processing Technology,
solvolthemal technology: a critical review,” Catalysts 2015, pp. 186–193.
[13] I. Sarantopoulos, E. Chatzisymeon, S. Foteinis, and T.
Reviews: Science and Engineering, 2015, pp. 1–29.
[2] A.P.S. Chouhan and A.K. Sarma, “Modern Tsoutsos, “Optimization of biodiesel production from
heterogeneous catalysts for biodiesel production: a waste lard by a two-step transesterification process
comprehensive review,” Renewable and Sustainable under mild conditions,” Energy for Sustainable
Energy Reviews, 2011, pp. 4378–4399. Development, 2014, pp. 110–114.
[3] Y.C. Sharma, B. Singh, and S.N. Upadhyay,
[14] H. Noureddini and D. Zhu, “Kinetics of
“Advancements in development and characterization transesterification of soybean oil,” Biocatalysis
of biodiesel: a review,” Fuel, 2008, pp. 2355–2373. Articles JAOCS, 1997, pp. 1457–1463.
[15] D.D. Steppan, J. Werner, and R.P. Yeater. Essential
[4] A. Srivastava and R. Prasad, “Triglyceride-based
diesel fuels,” Renewable and Sustainable Energy Regression and Experimental Design for Chemists
Reviews, 2000, pp. 111–133. and Engineers, 1998.
[16] C. Tongurai, Approximate analysis of total glyceride
[5] M.K. Lam, K.T. Lee, and A.R. Mohamed,
“Homogeneous, heterogeneous and enzymatic content in biodiesel by transesterification using
catalysis for transesterification of high free fatty acid microwave radiation technique. Thailand Patty Patent
oil (waste cooking oil) to biodiesel: A review,” No. 5060 August 18, 2008.
[17] M.L. Pisarello, B.O.D. Costa, N.S. Veizaga, and C.A.
Biotechnology Advances, 2010, pp. 500–518.
[6] G. Mendow, N.S. Veizaga, and C.A. Querini, “Ethyl Querini, “Volumetric method for free and total
ester production by homogeneous alkaline glycerin determination in biodiesel,” Ind. Eng. Chem.
transesterification: influence of the catalyst,” Res., 2010, pp. 8935–8941.
[18] J. Wu, J.L. Wang, M.H. Li, J.P. Lin, D.Z. Wei,
Bioresource Technology, 2011, pp. 6385–6391.
[7] J.M. Dias, M.C.M. Alvim-Ferraz, and M.F. Almeida, “Optimization of immobilization for selective
“Comparison of the performance of different oxidation of benzyl alcohol by Gluconobacter
homogeneous alkali catalysts during oxydans using response surface methodology,”
transesterification of waste and virgin oils and Bioresource Technology, 2010, pp. 8936–8941.
evaluation of biodiesel quality,” Fuel, 2008, pp.
[19] H. Hamze, M. Akia, F. Yazdani, “Optimization of
3572–3578. biodiesel production from the waste cooking oil using
[8] M.L. Pisarello and C.A. Querini, “Catalyst response surface methodology. Process Safety and
consumption during one and two steps Environmental Protection,” 2015, pp. 1–10.
transesterfication of crude soybean oils,” Chemical
[20] A. Bouaid, M. Martinez, J. Aracil, “A comparative
Engineering Journal, 2013, pp. 276–283. study of the production of ethyl esters from vegetable
[9] Z. Helwani, M.R. Othman, N. Aziz, J. Kim, and oils as a biodiesel fuel optimization by factorial
W.J.N. Fernando, “Solid heterogeneous catalyst for design,” Chem. Eng. Journal, 2007, pp. 93–99.
transesterification of triglycerides with methanol: A [21] X. Yuan, J. Liu, G. Zeng, J. Shi, J. Tong, G. Huang,
review,” Applied Catalyst A: General, 2009, pp. 1–10. “Optimization of conversion of waste rapeseed oil
[10] J. Li, Y.J. Fu, X.J. Qu, W. Wang, M. Luo, C.J. Zhao, with high FFA to biodiesel using response surface
and Y.G. Zu, “Biodiesel production from yellow horn methodology,” Renewable Energy, 2008, pp. 1678–
(Xanthoceras sorbifolia Bunge.) seed oil using ion 1684.
[22] D.C. Montgomery, Design and Analysis of
exchange resin as heterogeneous catalyst,”
Bioresource Technology, 2012, pp.112–118. Experiments, 5th ed. John Wiley and Sons, 2001.
[23] I. Noshadi, N.A.S. Amin, and R.S. Parnas,
[11] N. Boz, N. Degirmenbasi, and D.M. Kalyon,
“Esterification and transesterification of waste “Continuous production of biodiesel from waste
cooking oil over Amberlyst 15 and modified cooking oil in a reactive distillation column catalyzed
Amberlyst 15 catalysts,” Applied Catalysis B: by solid heteropolyacid: Optimization using response
Environmental, 2015, pp. 723–730. surface methodology (RSM),” Fuel 2012, pp. 156–
[12] Z.Z. Cai, Y. Wang, Y.L. Teng, K.M. Chong, J.W. 164.
[24] L.F. Bautista, G. Vicente, R. Rodriguez, and M.
 23rd Regional Symposium on Chemical Engineering (RSCE2016)
Innovation in Chemical Engineering towards the linkages among
education, academia, industry
27-28 Oct. 2016 in Vung Tau City, Vietnam

Pacheco, “Optimisation of FAME production from

waste cooking oil for biodiesel use,” Biomass and
Bioenergy, 2009, pp. 862–872.
[25] J.M. Avramovic, A.V. Velickovic, O.S. Stamenkovic,
K.M. Rajkovic, P.S. Milic, and V.B. Veljkovic,
“Optimization of sunflower oil ethanolysis catalyzed
by calcium oxide: RSM versus ANN-GA,” Energy
Conversion and Management, 2015, pp. 1149–1156.
[26] O.S. Stamenkovic, M.L. Lazic, Z.B. Todorovic, V.B.
VeljkovIC VB, and D.U. Skala, “The effect of
agitation intensity on alkali-catalyzed methanolysis of
sunflower oil,” Bioresource Technology, 2007, pp.
2688–2699.

View publication stats