A continuous-ow biodiesel production process using a rotating packed bed

Yi-Hung Chen
a,
*
, Yu-Hang Huang
b
, Rong-Hsien Lin
b
, Neng-Chou Shang
c
a
Department of Chemical Engineering and Biotechnology, National Taipei University of Technology, Taipei 106, Taiwan
b
Department of Chemical and Material Engineering, National Kaohsiung University of Applied Sciences, Kaohsiung 807, Taiwan
c
Department of Environmental Engineering, National I-Lan University, I-Lan 260, Taiwan
a r t i c l e i n f o
Article history:
Received 15 June 2009
Received in revised form 19 August 2009
Accepted 20 August 2009
Available online 13 September 2009
Keywords:
Micromixing
Hydraulic retention time
Fatty acid methyl esters
Productivity
Yield
a b s t r a c t
The continuous-ow transesterication of soybean oil with methanol using a rotating packed bed
(RPB) for the production of fatty acid methyl esters (biodiesel) is presented herein. The RPB, which
provides high centrifugal force and has an adjustable rotational speed, is employed as a novel transe-
sterication reactor. In this study, biodiesel is synthesized via the methanolysis of soybean oil using
potassium hydroxide as the catalyst. The following variables were investigated for their effects on
transesterication efciency: the methanoloil molar ratio, the estimated hydraulic retention time,
the rotational speed of the packed-bed rotator, the reaction temperature, and the catalyst dosage.
The yield of the fatty acid methyl esters (Y
FAME
) in the RPB system depends signicantly on the exper-
imental conditions, which inuence the residence time distribution, the transesterication reaction
rate, and the micromixing intensity. Due to its excellent micromixing characteristics, the RPB system
shows satisfactory transesterication efciency. The values of Y
FAME
, productivity of FAMEs (P
FAME
),
and P
FAME
per unit reactor volume (P
FAME
/V
R
) in the RPB are used to evaluate the performance for bio-
diesel production and allow for further comparison with other continuous transesterication reactors.
Consequently, a RPB is considered a practical transesterication reactor with high transesterication
efciency.
2009 Elsevier Ltd. All rights reserved.
1. Introduction
Biodiesel is usually composed of fatty acid methyl esters
(FAMEs) formed by the transesterication of triacylglycerols (oils
and fats) with methanol (methyl alcohol). This process is used to
reduce the viscosity and increase the volatility of the oils (Meher
et al., 2006a,b; Srivastava and Prasad, 2000). Recently, biodiesel
has drawn a great deal of attention because of the advantages asso-
ciated with its classication as a renewable energy resource, its
non-toxic character, and its biodegradability (Meher et al.,
2006b). Biodiesel contains no sulfur and has about 11% oxygen
by weight, and these properties lead to a remarkable reduction
in emissions compared with diesel fuel, especially unburned
hydrocarbons, sulfur dioxide, carbon monoxide, and particulate
matter (Demirbas, 2008; Sendzikiene et al., 2006).
The transesterication of triacylglycerols involves a number of
consecutive and reversible reactions. Alkali catalysts are frequently
used, including potassium hydroxide (KOH), sodium hydroxide
(NaOH), and the corresponding sodium or potassium methoxide
(KOCH
3
). The less-dense phase of the reactant mixture is the meth-
anol that contains the catalyst and the dense phase is made up of
triacylglycerols. The insolubility of methanol in the oil phase hin-
ders the progress of the transesterication reaction, so vigorous
mixing is essential to create sufcient contact between the two
immiscible phases (Ma et al., 1999; Srivastava and Prasad, 2000).
Biodiesel production in a continuous transesterication reactor is
usually limited by the retention time necessary to achieve the de-
sired degree of conversion of the triacylglycerols into fatty acid
methyl esters.
The conventional transesterication unit used in batch opera-
tion is considered slow, tedious, labor intensive, and not well-
adapted to automation. By comparison, a continuous-ow transe-
sterication process has advantages, including the adjustability
of the production cost and time for producing a small amount of
biodiesel, the ability to produce a larger biodiesel quantity per unit
of labor, and the capacity for making equipment design improve-
ments for the optimization of the biodiesel quality (Behzadi and
Farid, 2009). Accordingly, the use of a continuous-ow transesteri-
cation process saves time and money, thereby increasing produc-
tivity and prot, and contributing to the likelihood that biodiesel
eventually becomes a viable energy source.
A rotating packed bed (RPB) is designed to generate high cen-
trifugal acceleration, 13 orders of magnitude larger than gravita-
tional acceleration, via adjustments of the rotational speed
(Ramshaw and Mallinson, 1981). RPBs have been designed to
0960-8524/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2009.08.081
* Corresponding author. Tel.: +886 2 27712171x2539; fax: +886 2 87724328.
E-mail address: yhchen1@ntut.edu.tw (Y.-H. Chen).
Bioresource Technology 101 (2010) 668673
Contents lists available at ScienceDirect
Bioresource Technology
j our nal homepage: www. el sevi er . com/ l ocat e/ bi or t ech
increase the mass transfer and micromixing efciencies by using
great centrifugal force; this novel technology has been dubbed
Higee. RPBs were used as gasliquid contactors in absorption
(Lin et al., 2003; Chen et al., 2004, 2008c), distillation (Ramshaw,
1983; Kelleher and Fair, 1996), stripping (Singh et al., 1992; Liu
et al., 1996), and ozonation (Lin and Liu, 2003; Chen et al.,
2005a,b; Chang et al., 2009a,b), etc. As a result, RPBs demonstrate
high gasliquid mass transfer coefcients, which are crucial for
achieving high gasliquid mass transfer rates. Our previous study
(Chen et al., 2005b) also demonstrated that the injected liquid
would thoroughly disperse into a thin liquid lm, resulting in the
complete wetting of the packing. The mean liquid lm thickness
on the surface of the packing in the RPB was about 74182 lm,
which accounts for the high gasliquid mass transfer rate (Chen
et al., 2004).
Furthermore, RPBs have an excellent micromixing ability and
can be used for many novel applications in related industries (Lin
and Liu, 2003; Yang et al., 2005, 2006). For example, a RPB was re-
garded as an ideal reactor for the preparation of nanoparticles by a
reactive precipitation method because of the intensication of the
micromixing of uids (Chen et al., 2003). In comparison to other
reactors with regard to the production of nano-inorganic powders,
a RPB requires a shorter reaction time and produces smaller parti-
cles with a narrower size distribution range, thereby enabling eas-
ier industrialization (Zhao et al., 2007).
For the transesterication reaction, the reactants initially form
a two-phase liquid system. The reaction is diffusion-controlled
and the poor diffusion between the phases results in a slow reac-
tion rate (Srivastava and Prasad, 2000). The vigorous mixing
intensity is signicant in this reaction region. When the FAMEs
are generated, they act as a mutual solvent for the reactants
and a single-phase system is formed. As the single-phase is estab-
lished, the inuence of the mixing becomes insignicant to the
transesterication. On the other hand, the coagulation of glycerol
(by-product) starts to form the glycerol phase, which is positive
to the chemical equilibrium shift towards the transesterication
reaction. Note that the micromixing time in the RPB was evalu-
ated to be on the order of 10
4
s (Yang et al., 2005). For a RPB,
the residence time distribution (RTD) of liquid is within several
seconds (Chen et al., 2004). It is presumed that the reactants
would be intensively mixed within a very short time in a RPB
so the transesterication reaction could proceed quickly. Then
the great centrifugal force may contribute to the formation of
the glycerol phase.
The objective of this study is to develop a new technique for
the continuous production of biodiesel using a RPB as the transe-
sterication reactor, which represents a novel application of RPBs
in the eld of fuel generation. The methanolysis of soybean oil
with KOH catalyst is carried out in the RPB. One should notice
that the use of edible oils for biodiesel production may be against
current social movements and energy policies. Soybean oil, as one
common edible oil, is merely employed as a model oil and the ob-
tained results are considered to be relevant to other triacylglyce-
rols with different fatty acid compositions (Pinto et al., 2005).
Methanol is the preferred alcohol for obtaining biodiesel because
it is the cheapest alcohol (Van Gerpen et al., 2004). KOH is one of
the most common catalysts used for carrying out the transesteri-
cation (Meher et al., 2006b; Van Gerpen et al., 2004). The effects
of various experimental conditions on the yield of FAMEs (Y
FAME
,
%) are examined: the molar ratio of methanol to soybean oil (n
M
/
n
O
), the estimated hydraulic retention time (t
L
, min), the rota-
tional speed of the packed-bed rotator (x, rpm), the reaction tem-
perature (T, C), and the catalyst dosage based on oil weight (W
cat
,
% w/w). The value of Y
FAME
is dened as the percentage of the tri-
glyceride being transformed into biodiesel (He et al., 2005, 2006;
Wen et al., 2009).
2. Methods
2.1. Materials
Rened soybean oil was purchased from Uni-President Co., Ltd.
(Tainan, Taiwan) with a molecular weight (M
O
) of 875 g/mol; a
density (q
O
) of 0.917 g/ml; and an elemental composition of car-
bon (77.1 wt.%), hydrogen (10.4 wt.%), and oxygen (12.5 wt.%).
The elemental composition of the soybean oil was analyzed by a
F002 Heraeus CHNO Rapid Elemental Analyzer (Hanau, Ger-
many); sulfanilic acid and benzoic acid served as the standards.
The ACS-certied methanol obtained from Mallinckrodt Chemicals
(Phillipsburg, NJ, USA) has a molecular weight of 32 g/mol, a den-
sity of 0.79 g/ml, and a boiling point of 64.7 C. The KOH (85% pur-
ity) was purchased from Riedel-deHan (St. Gallen, Switzerland).
The FAME standards, which included palmitic acid, stearic acid,
oleic acid, linoleic acid, and linolenic acid methyl esters, were pur-
chased from Fluka (St. Gallen, Switzerland).
2.2. Instrumentation
The experimental apparatus employed in this work can be
found in detail in the patents (Chen et al., 2008a,b). The RPB con-
sisted of the packed-bed rotator and a stationary case; the speci-
cations of the RPB are given in Table 1. The reactor volume (V
R
) of
the RPB was estimated to be the volume of the packed-bed rotator,
pr
2
2
r
2
1
z
B
, in which r
1
, r
2
, and z
B
represent the inner radius, outer
radius, and axial height of the packed-bed rotator, respectively. The
rotator was connected to a rotor shaft on two bearings mounted to
a steel structure. The shaft was connected to a motor controlled by
a speed regulator. The value of x was in the range of 150
1500 rpm, which, based on the arithmetic mean radius of the
packed-bed rotator, provided a centrifugal acceleration of 0.94
94 gravitational acceleration (g
0
). Grade 304 stainless steel wires
cut in the shape of annular rings were stacked in the packed bed.
The density and diameter of the wires were 8478 kg/m
3
and
0.022 cm, respectively. Soybean oil and methanol were introduced
into the RPB via the liquid distributor, which had two vertical
groups of three holes per group, and the groups were spaced
180 apart. The liquids were sprayed on the inside edge of the rota-
tor and thrown outwards by the centrifugal force.
2.3. Experimental procedures
In all experiments, the temperature of the soybean oil and
methanol, which were kept in storage tanks, was controlled by
PID controllers at preset values. The PID controllers gain value
was optimized such that the heater kept the temperature within
1 C of the set point. Thermocouples were used to monitor the
temperature. A specic amount of KOH was added to the methanol
Table 1
Specications of the RPB reactor used in this study.
Item Unit Value
Inner radius of the packed-bed rotator (r
1
) cm 1.15
Outer radius of the packed-bed rotator (r
2
) cm 6.35
Axial height of the packed-bed rotator (z
B
) cm 2.0
Weight of packing in the packed-bed rotator g 91.24
Volume of the packed-bed rotator ml 245
Volume of the housing case ml 1780
Specic area of packing per unit volume of
the packed-bed rotator
cm
2
/cm
3
7.99
Total area of packing cm
2
1960
Void space in the packed-bed rotator cm
3
/cm
3
0.956
Rotational speed of the packed-bed rotator rpm 01500
Power consumption of the packed-bed rotator W 080
Y.-H. Chen et al. / Bioresource Technology 101 (2010) 668673 669
for the specic W
cat
value. The soybean oil and methanol contain-
ing KOH were simultaneously transported into the RPB at preset
ow rates to initiate transesterication. The tubing and RPB were
wrapped in glass ber heating tape (Morshine electric heat Co.,
Ltd., Taipei, Taiwan) to maintain the temperature of the liquids
during transportation. The t
L
was calculated as V
R
/(Q
OL
+ Q
ML
),
and was inversely proportional to the total ow rate where Q
OL
and Q
ML
were the ow rates of oil and methanol, respectively.
The stoichiometric value of the n
M
/n
O
ratio is 3 for transesteri-
cation. Nevertheless, the n
M
/n
O
ratio employed was often higher
than 3 to drive the equilibrium toward a higher Y
FAME
value, in
accordance with Le Chateliers principle (Ma and Hanna, 1999;
Meher et al., 2006b). In this study, the following transesterication
conditions were varied to evaluate their effects on Y
FAME
: n
M
/n
O
ra-
tio (4, 6, 12, 18, 24), t
L
(0.431.67 min), x(150, 300, 900, 1500 rpm),
T (34 C, 42 C, 60 C, 65 C), and W
cat
(0.1%, 0.25%, 1%, 2%, 3%, 5% w/
w). For each experimental run, the system was allowed to run long
enough to reach steady state; meanwhile, repeated samples were
taken from the efuent at the desired time intervals. The samples
were quenched by an ice batch and then left overnight to settle, a
process which resulted in phase separation. The cooling and gravity
separation procedures for the reaction mixture resembled the prac-
tical operation of biodiesel production (Barnard et al., 2007; Bouaid
et al., 2005; Dmytryshyn et al., 2004; Vicente et al., 2007), though
the use of mineral acid can stop the reaction immediately. Subse-
quently, the ester phase was washed with saturated salt solution
(three times the volume of the ester phase) to remove any residual
methanol, KOH, or glycerol that may be present.
The
1
H nuclear magnetic resonance (NMR) analyses of the bio-
diesel samples were performed on a Varian Unity INOVA 500 NMR
spectrometer (Varian Inc., Palo Alto, CA, USA) operating at
500 MHz and using CDCl
3
as the solvent. The Y
FAME
value was cal-
culated from the signal area ratio of the methoxy protons of the
methyl esters at 3.7 ppm to the methylene protons adjacent to
the ester group at 2.3 ppm in the
1
H NMR spectrum (Meher
et al., 2006b; Morgenstern et al., 2006; Suppes et al., 2001). The
accuracy of the Y
FAME
value obtained with NMR analysis was fur-
ther conrmed by comparison to HPLC analysis. The concentra-
tions of the FAMEs in the samples were analyzed in the HPLC
system model L-2000 series (Hitachi, Tokyo, Japan), equipped with
a silica-based column model STR ODS-II (5 lm, 250 4.6 mm I.D.)
(Shinwa Chemical Industries, Co., Ltd., Tokyo, Japan). The HPLC sys-
tem has a model L-2130 pump with an online vacuum degasser, a
model L-2200 autosampler with a variable injection capacity of
0.5500 ll, and a model L-2490 RI detector in series with a model
L-2455 diode array detector. EZChrom Elite software, which was
provided with the HPLC instrument, was used for the acquisition
and processing of data. A model Super co-150 column oven (En-
shine, Taipei, Taiwan) was used to maintain a constant column
temperature of 40 C. Acetonitrile was used at a ow rate of
1.0 ml/min for HPLC elution. In addition, the contents of glycerin,
and mono-, di- and triglycerides in the biodiesel sample that re-
lated to the conversion of acylglycerols were determined according
to EN 14105.
3. Results and discussion
The Y
FAME
and productivity of FAMEs (P
FAME
, in mol/min) values
of the continuous-ow transesterication process in the RPB sys-
tem are shown in Table 2; the value for P
FAME
is calculated as 3Q
OL-
q
O
Y
FAME
/M
O
. It is clear that the Y
FAME
and P
FAME
values are
signicantly dependent on the experimental conditions. The Y
FAME
value has a wide range (of 5.5097.3%), which is advantageous for
investigation of the characteristics of the continuous-ow transe-
sterication process in the RPB system. In this section, the effects
of n
M
/n
O
, t
L
, x, T, and W
cat
on the transesterication performance
are described, and the RPB is compared with other transesterica-
tion reactors.
3.1. Effect of n
M
/n
O
The variation of Y
FAME
with the n
M
/n
O
ratio in the RPB system is
shown in Table 2. The n
M
/n
O
ratio is adjusted by the Q
ML
, while the
Q
OL
is maintained at a constant 181 ml/min. When the n
M
/n
O
ratio
is elevated from 4 to 6, Y
FAME
increases from 58.6% to 64.8% as a re-
sult of the shift in the chemical equilibrium toward the products of
the transesterication reaction. However, Y
FAME
gradually de-
creases with high n
M
/n
O
ratios, specically when the n
M
/n
O
ratio
is above 6. One interpretation of these data is that a n
M
/n
O
ratio
higher than 6 slightly inuences the chemical equilibrium (Colucci
Table 2
Comparison of Y
FAME
and P
FAME
for methanolysis of soybean oil in RPB reactor under various experimental conditions.
Run no. Q
OL
(ml/min) Q
ML
(ml/min) n
M
/n
O
() t
L
(min) x (rpm) T (C) W
cat
(% w/w) Y
FAME
(%) P
FAME
(mol/min)
1 181 33 4 1.14 900 42 0.67 58.6 0.333
2 181 46 6 1.07 900 42 1 64.8 0.370
3 181 92 12 0.89 900 42 2 62.6 0.358
4 181 138 18 0.77 900 42 3 57.1 0.325
5 181 184 24 0.67 900 42 4 49.2 0.280
6 117 30 6 1.67 900 42 1 40.7 0.150
7 125 32 6 1.56 900 42 1 51.0 0.239
8 149 38 6 1.31 900 42 1 58.1 0.319
9 271 69 6 0.72 900 42 1 65.4 0.557
10 362 92 6 0.54 900 42 1 61.4 0.698
11 450 115 6 0.43 900 42 1 49.2 0.695
12 181 46 6 1.07 150 42 1 19.0 0.109
13 181 46 6 1.07 300 42 1 26.9 0.154
14 181 46 6 1.08 1500 42 1 58.3 0.331
15 181 46 6 1.08 900 42 0.1 5.50 0.031
16 181 46 6 1.08 900 42 0.25 16.7 0.094
17 181 46 6 1.08 900 42 2 85.0 0.480
18 181 46 6 1.08 900 42 3 94.0 0.534
19 181 46 6 1.08 900 42 5 93.7 0.533
20 181 46 6 1.08 900 34 1 37.8 0.215
21 181 46 6 1.07 900 60 1 94.0 0.537
22 181 46 6 1.07 900 65 1 80.4 0.457
23 181 46 6 1.07 900 60 2 95.2 0.544
24 181 46 6 1.07 900 60 3 96.3 0.550
25 271 69 6 0.72 900 60 3 97.3 0.828
670 Y.-H. Chen et al. / Bioresource Technology 101 (2010) 668673
et al., 2005; Noureddini et al., 1998; Ma et al., 1999; Meher et al.,
2006a). On the other hand, excess methanol, with its afnity for
the hydrophilic stainless steel wire, may change the ow pattern
of the liquids (Boulang-Petermann et al., 2006). Thus, the holdup
volume of the oily liquids in the packed-bed rotator would de-
crease, resulting in shorter RTD. Therefore, the optimal n
M
/n
O
ratio
for transesterication in the RPB system is about 6.
3.2. Effect of t
L
As shown in Fig. 1, the effect of the t
L
on Y
FAME
and P
FAME
was
studied by adjusting the values of both Q
OL
and Q
ML
and maintain-
ing the n
M
/n
O
ratio at a constant of 6. Smaller t
L
value means an in-
crease in the liquid ow rate. For a specic liquid distributor, a
higher liquid ow rate enhances the micromixing efciency (Yang
et al., 2005). Accordingly, the Y
FAME
value increased, from 40.7% to
65.4%, with the decrease of the t
L
, from 1.67 to 0.72 min because
of more intensive micromixing. Nevertheless, Y
FAME
started to de-
crease when t
L
decreased to a value smaller than 0.72 min. This is
because the RTD of the injected liquid remarkably decreased with
the increase of the liquid ow rate when the t
L
was smaller than
0.72 min (Chen et al., 2004; Yang et al., 2005, 2006), thus leading
to a decrease in Y
FAME
, while the micromixing efciency varied
slightly. Furthermore, since P
FAME
is calculated as 3Q
OL
q
O
Y
FAME
/
M
O
, it increased with smaller t
L
values, owing to the larger product
of the Q
OL
and Y
FAME
values. When t
L
is below 0.54 min, P
FAME
ap-
proached a steady maximum value of about 0.70 mol/min as a con-
sequence of simultaneously increasing Q
OL
and decreasing Y
FAME
.
3.3. Effect of x
Table 2 depicts the variation of Y
FAME
with the x which repre-
sents the effect of the centrifugal acceleration on the transesteri-
cation performance in the RPB. The higher centrifugal acceleration
enhances the relative velocity amongst the liquids and packing and
reduces the thickness of the liquid lm on the packing based on the
balance of inertial and centrifugal forces (Chen et al., 2004). The va-
lue of Y
FAME
increased markedly when xwas increased from 150 to
900 rpm, because of the intensied micromixing between the oil
and methanol phases. In contrast, Y
FAME
appears to decrease when
x is elevated to 1500 rpm. That implies that only an insignicant
enhancement in the micromixing efciency occurs when x is
raised from 900 to 1500 rpm. Instead, the RTD of the liquids in
the RPB decreased with higher x because of the decreased liquid
holdup (Chen et al., 2004; Yang et al., 2005), which accounts for
the decrease in Y
FAME
at 1500 rpm. Therefore, the optimal x value
for transesterication in the RPB system is 900 rpm, which pro-
vides 1057 g
0
inside the packed-bed rotator.
3.4. Effect of T
The marked effect of the T on Y
FAME
in the RPB system is illus-
trated in Table 2. The value of Y
FAME
has distinct values of 37.8%,
64.8%, and 94.0% at T values of 34 C, 42 C, and 60 C, respectively.
It is clear that the transesterication reaction rate, which signi-
cantly increases with T (Ma and Hanna, 1999), affects Y
FAME
in
the RPB system, especially due to the RPB systems short RTD. A
higher T should be adopted to achieve a higher Y
FAME
in a RPB. Cer-
tainly, the value of T should be limited to the boiling point of meth-
anol to prevent the adverse effect of vaporization, which was
demonstrated by the decreasing Y
FAME
of 80.4% at T of 65 C. Also,
a negative effect on Y
FAME
may arise from the increasing solubility
of glycerol in FAMEs at higher temperature.
3.5. Effect of W
cat
The variations in Y
FAME
with changes in the W
cat
at T = 42 C and
60 C are shown in Table 2. At the lower T of 42 C, the value of
Y
FAME
apparently increases with higher W
cat
, a trend that can be
Fig. 1. Effect of t
L
on Y
FAME
and P
FAME
for methanolysis of soybean oil in the
continuous-ow RPB reactor. n
M
/n
O
= 6, x= 900 rpm, W
cat
= 1% w/w, and T = 42 C.
s and 4: Y
FAME
and P
FAME
.
Table 3
The Y
FAME
, P
FAME
, and P
FAME
/V
R
values obtained using various continuous transesterication reactors (as found in the literature) compared with those obtained using the present
RPB system.
a
Author Oil Catalyst Reactor T
(C)
V
R
(ml) t
L
(min)
Y
FAME
(%) P
FAME
(mol/min) P
FAME
/V
R
10
3
(mol/(ml min))
Noureddini et al.
(1998)
Soybean oil NaOH Motionless mixers and
high-shear reactor
80 2000 5.80
8.12
6899.5 0.5110.872 0.2550.436
Darnoko and
Cheryan (2000)
Palm oil KOH Stirred-tank reactor 60 1000 4070 58.897.3 0.02520.0278 0.02520.0278
He et al. (2005) Canola oil KOH,
KOCH
3
Reactive distillation reactor 55 10 2.67
6.67
41.597.0 0.00172
0.00931
0.1720.931
He et al. (2006) Canola oil KOH Reactive distillation reactor 65 10 2 61.994.4 0.01270.0158 1.271.58
Barnard et al.
(2007)
Soybean oil KOH Microwave heating reactor 50 2000,
4000
0.56
2
94.495.25 4.9017.8 1.234.46
Stavarache et al.
(2007)
Commercial oil,
palm oil
KOH Ultrasonication 38
40
2620,
6350
1030 72.096.0,
50.094.0
0.1990.623,
0.1860.514
0.0570.23, 0.058
0.119
Wen et al. (2009) Soybean oil NaOH Zigzag micro-channel
reactor
40
75
0.0642 0.3,
0.47
81.599.5 0.00025
0.00052
3.898.11
Present work Soybean oil KOH RPB 34
65
245 0.43
1.67
5.5097.3 0.0310.828 0.1273.378
a
The alcohol used in the transesterication reaction is methanol.
Y.-H. Chen et al. / Bioresource Technology 101 (2010) 668673 671
attributed to the acceleration of the transesterication reaction
rate by higher methoxide anion concentration. In comparison,
the effect of W
cat
on Y
FAME
weakens at the higher T of 60 C, when
the transesterication reaction rate is already high. The W
cat
value
of 1% w/w has been considered practicable for the common transe-
sterication process. When T is low, an increase in W
cat
can signif-
icantly improve the transesterication performance in the RPB
system. However, one should note that Y
FAME
slightly varies in
the W
cat
region of 35% w/w.
3.6. Comparison with other continuous transesterication reactors
Based on the results, Run No. 25 in Table 2 represents optimal
conditions, as that run yielded the highest Y
FAME
(97.3%) and
P
FAME
(0.828 mol/min) values. Further the contents of mono-, di-
and triglyceride, and free and total glycerine are determined as
0.093%, 0.014%, 0%, 0%, and 0.107% (w/w), respectively, which
meet the requirements dened in EN 14214 standard. In Table
3, the experimental conditions and transesterication perfor-
mance of the RPB are compared with those of other continuous
transesterication reactors. Among the continuous transesterica-
tion reactors, the RPB reactor has a small t
L
and a high Y
FAME
,
P
FAME
, and P
FAME
per unit reactor volume (P
FAME
/V
R
, mol/(ml min))
values. The small t
L
implies that short operation time is required
to reach steady state since the continuous transesterication sys-
tem is usually allowed to stabilize for about 3 t
L
(Noureddini
et al., 1998). The satisfactory Y
FAME
, P
FAME
, and P
FAME
/V
R
values
of the RPB system reveal its unique micromixing ability. A RPB
also has the advantages of low cost, simple operation, and easy
maintenance. Consequently, a RPB has high potential to be an
effective continuous transesterication reactor for biodiesel
production.
4. Conclusions
With respect to the yield of fatty acid methyl esters (Y
FAME
), the
optimal molar ratio of methanol to oil, estimated hydraulic reten-
tion time, and rotational speed in the rotating packed bed (RPB)
system are 6, 0.72 min, and 900 rpm, respectively. The higher reac-
tion temperature and catalyst dosage are advantageous for Y
FAME
.
The present RPB system achieved a maximum Y
FAME
value of
97.3% and maximum productivity of fatty acid methyl esters
(P
FAME
) of 0.828 mol/min. Compared to other continuous transeste-
rication reactors, the RPB has a small estimated hydraulic reten-
tion time and high Y
FAME
, P
FAME
, and P
FAME
per reactor volume
(P
FAME
/V
R
).
Acknowledgements
This study was supported by the National Science Council of
Taiwan under Grant Nos. NSC 95-2221-E-151-025, NSC 96-2221-
E-027-140 -MY2, and NSC 97-2621-M-027-005.
References
Barnard, T.M., Nicholas, E.L., Matthew, B.B., Lauren, M.S., Benjamin, A.W., 2007.
Continuous-ow preparation of biodiesel using microwave heating. Energ. Fuel.
21, 17771781.
Behzadi, S., Farid, M.M., 2009. Production of biodiesel using a continuous gasliquid
reactor. Bioresource Technol. 100, 683689.
Bouaid, A., Diaz, Y., Martinez, M., Aracil, J., 2005. Pilot plant studies of biodiesel
production using Brassica carinata as raw material. Catal. Today 106, 193196.
Boulang-Petermann, L., Gabeta, C., Baroux, B., 2006. On the respective effect of the
surface energy and micro-geometry in the cleaning ability of bare and coated
steels. Colloid. Surf. A 272, 5662.
Chang, C.C., Chiu, C.Y., Chang, C.Y., Chang, C.F., Chen, Y.H., Ji, D.R., Tseng, J.Y., Yu, Y.H.,
2009a. Pt-catalyzed ozonation of aqueous phenol solution using high-gravity
rotating packed bed. J. Hazard. Mater. 168, 649655.
Chang, C.C., Chiu, C.Y., Chang, C.Y., Chang, C.F., Chen, Y.H., Ji, D.R., Yu, Y.H., Chiang,
P.C., 2009b. Combined photolysis and catalytic ozonation of dimethyl phthalate
in a high-gravity rotating packed bed. J. Hazard. Mater. 161, 287293.
Chen, J.F., Shao, L., Guo, F., Wang, X.M., 2003. Synthesis of nano-bers of
aluminum hydroxide in novel rotating packed bed reactor. Chem. Eng. Sci. 58,
569575.
Chen, Y.H., Chang, C.Y., Lin, R.H., Huang, Y.H., 2008a. Method of increasing
transesterication of oils. Patent No. TW200831660-A.
Chen, Y.H., Chang, C.Y., Lin, R.H., Huang, Y.H., 2008b. Method of increasing
transesterication of oils. Patent No. US2008171889-A1.
Chen, Y.H., Chang, C.Y., Su, W.L., Chen, C.C., Chiu, C.Y., Yu, Y.H., Chiang, P.C., Chiang,
Sally I.M., 2004. Modeling ozone contacting process in a rotating packed bed.
Ind. Eng. Chem. Res. 43, 228236.
Chen, Y.H., Chang, C.Y., Su, W.L., Chiu, C.Y., Yu, Y.H., Chiang, P.C., Chang, C.F., Shie,
J.L., Chiou, C.S., Sally, I.M., 2005a. Ozonation of CI reactive black 5 using
rotating packed bed and stirred tank reactor. J. Chem. Technol. Biotechnol. 80,
6875.
Chen, Y.H., Chiu, C.Y., Chang, C.Y., Huang, Y.H., Yu, Y.H., Chiang, P.C., Shie, J.L., Chiou,
C.S., 2005b. Modeling ozonation process with pollutant in a rotating packed
bed. Ind. Eng. Chem. Res. 44, 2129.
Chen, Y.S., Hsu, Y.C., Lin, C.C., Tai, C.Y., Clifford, Y.D., Liu, H.S., 2008c. Volatile organic
compounds absorption in a cross-ow rotating packed bed. Environ. Sci.
Technol. 42, 26312636.
Colucci, J.A., Borrero, E.E., Alape, F., 2005. Biodiesel from an alkaline
transesterication reaction of soybean oil using ultrasonic mixing. J. Am. Oil
Chem. Soc. 82, 525530.
Darnoko, D., Cheryan, M., 2000. Continuous production of palm methyl esters. J. Am.
Oil Chem. Soc. 77, 12691272.
Demirbas, A., 2008. Biodiesel: A Realistic Fuel Alternative for Diesel Engines, rst ed.
Springer-Verlag, London.
Dmytryshyn, S.L., Dalai, A.K., Chaudhari, S.T., Mishra, H.K., Reaney, M.J., 2004.
Synthesis and characterization of vegetable oil derived esters: evaluation for
their diesel additive properties. Bioresource Technol. 92, 5564.
He, B.B., Singh, A.P., Thompson, J.C., 2005. Experimental optimization of a
continuous-ow reactive distillation reactor for biodiesel production. Trans.
ASAE 48, 22372243.
He, B.B., Singh, A.P., Thompson, J.C., 2006. A novel continuous-ow reactor using
reactive distillation for biodiesel production. Trans. ASAE 49, 107112.
Kelleher, Y., Fair, J.R., 1996. Distillation studies in a high-gravity contactor. Ind. Eng.
Chem. Res. 12, 46464655.
Lin, C.C., Liu, W.T., 2003. Ozone oxidation in a rotating packed bed. J. Chem. Tech.
Biotechnol. 78, 138141.
Lin, C.C., Liu, W.T., Tan, C.S., 2003. Removal of carbon dioxide by absorption in a
rotating packed bed. Ind. Eng. Chem. Res. 11, 23812386.
Liu, H.S., Lin, C.C., Wu, S.C., Hsu, H.W., 1996. Characteristic of a rotating packed bed.
Ind. Eng. Chem. Res. 35, 25963590.
Ma, F., Clement, L.D., Hanna, M.A., 1999. The effect of mixing on transesterication
of beef tallow. Bioresource Technol. 69, 289293.
Ma, F., Hanna, M.A., 1999. Biodiesel production: a review. Bioresource Technol. 70,
115.
Meher, L.C., Dharmagadda, V.S.S., Naik, S.N., 2006a. Optimization of alkali-catalyzed
transesterication of Pongamia pinnata oil for production of biodiesel.
Bioresource Technol. 97, 13921397.
Meher, L.C., Sagar, D.V., Naik, S.N., 2006b. Technical aspects of biodiesel production
by transesterication a review. Renew. Sust. Energ. Rev. 10, 248268.
Morgenstern, M., Cline, J., Meyer, S., Cataldo, S., 2006. Determination of the kinetics
of biodiesel production using proton nuclear magnetic resonance spectroscopy
(
1
H NMR). Energ. Fuel. 20, 13501353.
Noureddini, H., Harkey, D., Medikonduru, V., 1998. A continuous process for the
conversion of vegetable oils into methyl esters of fatty acids. J. Am. Oil Chem.
Soc. 75, 17751783.
Pinto, A.C., Guarieiro, L.L.N., Rezende, M.J.C., Ribeiro, N.M., Torres, E.A., Lopes, W.A.,
de P. Pereira, P.A., de Andrade, J.B., 2005. Biodiesel: an overview. J. Brazil Chem.
Soc. 16, 13131330.
Ramshaw, C., 1983. Higee distillation an example of process intensication. Chem.
Eng. 2, 1314.
Ramshaw, C., Mallinson, R.H., 1981. Mass transfer process. US Patent 4,283,255.
Sendzikiene, E., Makareviciene, V., Janulis, P., 2006. Inuence of fuel oxygen content
on diesel engine exhaust emissions. Renew. Energ. 31, 25052512.
Singh, S.P., Wilson, J.H., Counce, R.M., Villiers-Fisher, J.F., Jennings, H.L., Lucero, A.J.,
Reed, G.D., Ashworth, R.A., Elliott, M.G., 1992. Removal of volatile organic
compounds from groundwater using a rotary air stripper. Ind. Eng. Chem. Res.
31, 574580.
Srivastava, A., Prasad, R., 2000. Triglycerides-based diesel fuels. Renew. Sust. Energ.
Rev. 4, 111133.
Stavarache, C., Vinatoru, M., Maeda, Y., Bandow, H., 2007. Ultrasonically driven
continuous process for vegetable oil transesterication. Ultrason. Sonochem.
14, 413417.
Suppes, G.J., Bockwinkel, K., Lucas, S., Botts, J.B., Mason, M.H., Heppert, J.A., 2001.
Calcium carbonate catalyzed alcoholysis of fats and oils. J. Am. Oil Chem. Soc.
78, 139145.
Van Gerpen, J., Shanks, B., Pruszko, R., Clements, D., Knothe, G., 2004. Biodiesel
production technology. NREL/SR-510-36244, Golden, Colorado.
Vicente, G., Martnez, M., Aracil, J., 2007. Optimisation of integrated biodiesel
production. Part I. A study of the biodiesel purity and yield. Bioresource
Technol. 98, 17241733.
672 Y.-H. Chen et al. / Bioresource Technology 101 (2010) 668673
Wen, Z., Yu, X., Tu, S.T., Yan, J., Dahlquist, E., 2009. Intensication of biodiesel
synthesis using zigzag micro-channel reactors. Bioresource Technol. 100, 3054
3060.
Yang, H.J., Chu, G.W., Xiang, Y., Chen, J.F., 2006. Characterization of micromixing
efciency in rotating packed beds by chemical methods. Chem. Eng. J. 121, 147
152.
Yang, H.J., Chu, G.W., Zhang, J.W., Shen, Z.G., Chen, J.F., 2005. Micromixing efciency
in a rotating packed bed: experiments and simulation. Ind. Eng. Chem. Res. 44,
77307737.
Zhao, R.H., Li, C.P., Guo, F., Chen, J.F., 2007. Scale-up preparation of organized
mesoporous alumina in a rotating packed bed. Ind. Eng. Chem. Res. 46, 3317
3320.
Y.-H. Chen et al. / Bioresource Technology 101 (2010) 668673 673