Professional Documents
Culture Documents
Renewable Energy
journal homepage: www.elsevier.com/locate/renene
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 October 2013
Accepted 3 March 2014
Available online
The present study deals with the production of biodiesel using waste sh oil. The research assesses the
effect of the transesterication parameters on the biodiesel yield and its properties, including temperature (40e60 C), molar ratio methanol to oil (3:1e9:1) and reaction time (30e90 min). The experimental results were tted to complete quadratic models and optimized by response surface
methodology. All the biodiesel samples presented a FAME content higher than 93 wt.% with a maximum,
95.39 wt.%, at 60 C, 9:1 of methanol to oil ratio and 90 min. On the other hand, a maximum biodiesel
yield was found at the same methanol to oil ratio and reaction time conditions but at lower temperature,
40 C, which reduced the saponication of triglycerides by the alkaline catalyst employed. Adequate
values of kinematic viscosity (measured at 30 C) were obtained, with a minimum of 6.30 mm2/s obtained at 60 C, 5.15:1 of methanol to oil ratio and 55.52 min. However, the oxidative stability of the
biodiesels produced must be further improved by adding antioxidants because low values of IP, below
2.22 h, were obtained. Finally, satisfactory values of completion of melt onset temperature, ranging from
3.31 C to 3.83 C, were measured.
2014 Elsevier Ltd. All rights reserved.
Keywords:
Fish waste oil
Biodiesel
Alkaline transesterication
Kinematic viscosity
Oxidative stability
Completion of melt onset temperature
1. Introduction
The sh industry still generates large quantities of by-products,
mainly because of sh processing and discards [1]. Although these
wastes are generally employed to obtain low value products such as
sh silage or fertilizer, highly valued commodities such as sh oil
can be also produced from them [2,3]. Fish oil containing a high
content of omega-3 PUFA (>20 wt.%) has gained an increasing interest in the pharmaceutical and functional food industries, due to
the recognized health benets of eicosapentaenoic (C20:5n-3) and
docosahexaenoic acids (C22:6n-3) [4,5]. However, sh oil having a
lower amount of EPA and DHA leads to very low yield of omega-3
concentrates which considerably reduces its applications on both
elds [6]. In addition, sh oil extracted from sh wastes may not
meet the quality criteria required for edible purposes [7].
In this framework, alternative uses for poor quality sh oil, for
example as a fuel, need to be investigated. Although crude sh oil
can be directly employed as combustible in boilers or furnaces, it
may results in large particulate emissions and carbon deposits on
619
(1)
Yield
MBiodiesel C
100
MOil
(2)
620
Y b0
3
X
i1
bi X i
3
X
i1
bii Xi2
3
X
bij Xi Xj
(3)
Table 1
Experimental design and measured values for the response variables.
Exp#
Temp.
( C)
Methanol:Oil
ratio
Time
(min)
FAME
(wt.%)
Yield
(%)
Viscosity
(mm2/s)
IP (h)
COM
( C)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
40
40
40
40
40
40
40
40
40
50
50
50
50
50
50
50
50
50
60
60
60
60
60
60
60
60
60
3
3
3
6
6
6
9
9
9
3
3
3
6
6
6
9
9
9
3
3
3
6
6
6
9
9
9
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
93.12
93.26
93.72
93.98
93.65
93.76
93.58
93.35
93.53
93.08
93.27
93.43
93.87
93.91
94.19
93.60
93.59
95.14
94.19
93.95
94.65
94.05
94.10
93.56
93.15
94.78
95.39
74.49
76.14
72.16
78.94
80.16
80.02
83.06
77.71
82.40
71.67
72.47
76.19
76.50
73.25
80.77
74.51
78.24
81.01
79.31
76.43
72.31
77.83
71.19
72.97
79.92
79.95
75.22
6.66
6.45
6.41
6.48
6.45
6.57
6.51
6.35
6.56
6.61
6.38
6.46
6.33
6.40
6.42
6.37
6.43
6.43
6.37
6.39
6.40
6.45
6.05
6.51
6.40
6.49
6.62
1.29
1.43
1.29
1.21
1.30
1.27
1.10
1.31
1.36
1.80
1.05
1.25
1.40
1.21
1.38
1.53
1.44
1.27
1.78
1.86
1.21
1.83
1.16
1.20
2.22
1.50
1.23
3.70
3.71
3.72
3.70
3.75
3.83
3.76
3.79
3.82
3.70
3.73
3.55
3.69
3.75
3.48
3.73
3.82
3.58
3.72
3.74
3.44
3.54
3.74
3.37
3.31
3.69
3.77
i<j
where the coefcients bi and bii are related to the linear and
quadratic effects, respectively, of each input factor on the output
variable and the cross-product coefcients bij represent the interactions between two input variables.
Secondly, the analysis of variance (ANOVA) was carried out. The
signicance of all terms in the models was judged statistically by
computing the p-value at a condence level of 95%. The regression
coefcients were then used to generate contour maps and to nd
the processing conditions which lead to an optimum (maximum or
minimum) value for a given response variable. The optimum solution should be obtained by a combination of the experimental
factors inside their ranges of application.
3. Results and discussion
3.1. Characterization of biodiesel samples
Table 1 shows the experimental values of FAME content, yield,
kinematic viscosity, IP and COM for the biodiesels produced.
3.1.1. FAME content and yield
The biodiesel purity was assayed by determining its FAME
content. A good conversion of triglycerides to FAME was obtained
for all the samples because their FAME content were higher than
93 wt.% (Table 1). Although the inuence of the processing conditions was not clear, it was seen that when maintaining constant the
methanol to oil ratio at 9:1 and the reaction time at 90 min, the
FAME content increased with temperature (from 93.53 wt.% at
40 C to 95.39 wt.% at 60 C). Besides, the FAME content improved
at larger reaction time when employing a methanol to oil ratio of
9:1 at 50 C and 60 C but not at 40 C. Hence, the maximum
conversion, 95.39 wt.%, was obtained at 60 C with a 9:1 methanol
to oil ratio and a reaction time of 90 min. Nevertheless, it should be
mentioned that this value is slightly lower than the minimum
required by the European standard (EN 14214) for use as 100%
diesel fuel (96.5 wt.%).
The fatty acids of the biodiesel samples were composed primarily of 28.7 wt.% oleic acid (C18:1n-9), 11.8 wt.% palmitic acid
(C16:0), 10.2 wt.% linoleic acid (C18:2n-6), 7.6 wt.% DHA (C22:6n-3),
5.7 wt.% EPA (C20:5n-3) and 5.0 wt.% palmitoleic acid (C16:1n-7).
The content of saturated fatty acids (SFA) and polyunsaturated fatty
acids (PUFA) amounted to 17.7 wt.% and 33.2 wt.%, respectively. This
was in agreement with the SFA content of waste cooking oil biodiesel (19.8 wt.%) and with the PUFA content of marine oil biodiesel
(37.3 wt.%) [18].
The biodiesel yields obtained ranged from 71.19 to 83.06%
(Table 1). It was observed that the highest percentage yields were
found at low temperature (40 C) and high methanol to oil ratio
(9:1). Although the reaction time had a positive effect on yield at
50 C, since at constant methanol to oil ratio the yield increases
with longer contact time, this trend was not seen at 40 C or at
60 C. These yield values are in line with those reported for biodiesel produced with oil extracted from sh canning industry
wastes, 73.9% using 1 wt.% of acid catalyst [19] and with menhaden
oil, 86.2% employing 2 wt.% of alkaline catalyst [20]. However, they
were considerably lower than those found for biodiesel obtained
with salmon oil, 99% [7] and with heckel sh oil, 96% [21]. It may be
due to the fact that the latter two studies employed a lower concentration of alkaline catalyst, 0.5 wt.%. According to Fadhil and Ali
[21], an alkaline catalyst (NaOH or KOH) concentration beyond
0.5 wt.% enhanced the formation of soaps by triglyceride saponication. Owing to their polarity, soaps are dissolved into the glycerol phase increasing also the methyl ester solubility in the glycerol,
which is a cause of yield loss [22].
3.1.2. Viscosity
Viscosity is an important fuel property since it affects the injection and atomization of a fuel. The higher the viscosity, the
greater the tendency of a fuel to cause operational problems such as
engine deposits [23]. The kinematic viscosity of the biodiesel
samples varied between 6.05 and 6.66 mm2/s at 30 C (Table 1).
Thus, the transesterication process was effective at reducing the
621
unsaturated fatty acids. The last visible peak was obtained between
20 C and 3.83 C, which may mainly consist of methyl esters of
SFA. Hence, the melting point of the biodiesel samples ranged from
73.60 C to 3.83 C. Similar endothermic peaks were found by
Garcia-Prez et al. [28] for biodiesel produced from poultry fat,
with a melting point between 69.5 C and 14.2 C. Subsequently,
DSC curves were analyzed to determine the COM for all the biodiesel samples (Table 1). COM is dened as the temperature at
which the melting phenomenon nishes and it ranged from 3.31 C
to 3.83 C. This small variation is due to the fact that transesterication does not alter the fatty acid composition of the
feedstocks and thus, it is the fatty acid prole of the raw oil which
determines the melting properties of the biodiesel [24]. A similar
value was found by Chiou et al. [10] for salmon oil methyl esters,
3.1 C; but a lower one was reported by the same authors for corn
oil methyl esters, 1.7 C, which is explained by its lower saturated
fatty acids content.
3.1.5. Acidity
The acid value of biodiesel is considered as a quality control
parameter given that it determines its corrosiveness and long-term
stability [29]. In this study, all the biodiesel samples showed an acid
value below 0.31 mg KOH/g oil after the transesterication step,
which is in accordance with the European standard (<0.5 mg KOH/
g oil).
3.2. Statistical modeling
The experimental data for each output variable were tted to a
complete second order model. Table 2 shows the polynomial coefcients calculated by multiple regressions for each response
variable and their associated p-value. It was observed that the
linear effect of temperature was statistically signicant for FAME
content, IP and COM. The linear effect of reaction time was also
signicant for FAME content and IP. On the other hand, the linear
effect of methanol to oil ratio was found to be statistically signicant only for yield. Quadratic effects were signicant only for reaction time in the case of viscosity and COM. The p-values for the
remaining effects indicated that only the interaction between
temperature and reaction time was statistically signicant in the
case of IP. Nevertheless, it should be mentioned that the proposed
quadratic models seems to not explain the variability of the data to
a large extent. The highest coefcient of determination was obtained for IP, R2 0.7076, whereas values above 0.5 were found for
the rest of the output variables (Table 2).
3.3. Optimization by response surface methodology
For the optimization of the biodiesel production, the yield was
chosen as measured variable showing the efciency of the process,
whereas viscosity, IP and COM were selected as variables indicating
the fuel quality of the biodiesel produced. The goodness of the t
for these variables was also proved by plotting the measured values
against the predicted ones for yield (Fig. 2a), viscosity (Fig. 2b), IP
(Fig. 2c) and COM (Fig. 2d). The data were correlated by means of a
regression line whose equation was inserted in each gure.
Although a low coincidence between the regression line and the
diagonal was observed for the four variables, most of the experimental points were placed inside the region delimited by the
dotted lines at both sides of the diagonal, which represent a deviation of 10% between predicted and measured values. Subsequently, by means of the second-order models obtained above and
using response surface methodology, contour maps were generated
where the yield, viscosity, IP and COM were plotted against time of
reaction and methanol to oil molar ratio, as shown in Fig. 3aed.
622
Table 2
Polynomial coefcients and p-values for the response variables.
FAME (wt.%)
Yield (%)
Viscosity (mm2/s)
IP (h)
COM ( C)
Coefcient
p-Value
Coefcient
p-Value
Coefcient
p-Value
Coefcient
p-Value
Coefcient
p-Value
Constant
A: Temperature, C
B: Methanol:Oil ratio
C: Time, min
AA
AB
AC
BB
BC
CC
9.61E01
8.88E02
1.39E02
4.70E02
8.78E04
4.26E04
5.19E04
8.33E03
2.15E03
1.41E04
e
0.0139
0.1249
0.0415
0.6760
0.9310
0.3007
0.7207
0.2025
0.5478
8.86E01
8.10E01
2.28
2.97E02
1.17E02
3.70E02
4.07E03
4.80E04
5.51E03
1.12E03
e
0.1371
0.0053
0.8066
0.3565
0.2185
0.1785
0.9975
0.5762
0.4226
8.65
4.20E02
1.72E01
2.23E02
2.33E04
1.24E03
1.16E04
5.96E03
6.43E04
1.08E04
e
0.0784
0.9442
0.6527
0.5692
0.2067
0.2370
0.1985
0.0569
0.0268
9.86E01
2.53E03
2.02E01
2.09E02
4.69E04
9.44E04
6.94E04
1.23E02
1.11E04
6.79E05
e
0.0061
1.0000
0.0029
0.5172
0.5989
0.0010
0.1555
0.8401
0.4250
3.65
8.33E04
3.01E02
1.29E02
2.22E05
1.03E03
5.56E05
3.58E03
7.22E04
1.25E04
e
0.0072
0.5938
0.5523
0.9621
0.3567
0.6152
0.4933
0.0621
0.0254
R2
0.5221
0.5171
0.5294
0.7076
0.5599
Fig. 2. Correlation between predicted and measured values of (a) Yield, (b) Viscosity, (c) IP and (d) COM.
623
Fig. 3. Contour plots for (a) Yield at 40 C, (b) Viscosity, (c) IP and (d) COM at 60 C.
624
[15]
References
[16]
[1] Rustad T. Physical and chemical properties of protein seafood by-products. In:
Shahidi F, editor. Maximising the value of marine by-products. Cambridge:
Woodhead Publishing Limited; 2006. pp. 3e21.
[2] Aidos I, Kreb N, Boonman M, Luten JB, Boom RM, van der Padt A. Inuence of
production process parameters on sh oil quality in a pilot plant. J Food Sci
2003;68:581e7.
[3] Wu TH, Bechtel PJ. Salmon by-product storage and oil extraction. Food Chem
2008;111:868e71.
[4] Uauy T, Valenzuela A. Marine Oils: the health benets of n-3 fatty acids.
Nutrition 2000;16:680e4.
[5] Ward OP, Singh A. Omega-3/6 fatty acids: alternative sources of production.
Process Biochem 2005;40:3627e52.
[6] Bimbo AP. Current and future sources of raw materials for the long-chain
omega-3 fatty acid market. Lipid Technol 2007;19:176e9.
[7] El-Mashad HM, Zhang R, Avena-Bustillos RJ. A two-step process for biodiesel
production from salmon oil. Biosyst Eng 2008;99:220e7.
[8] Ma F, Hanna MA. Biodiesel production: a review. Bioresour Technol 1999;70:
1e15.
[9] Helwani Z, Othman MR, Aziz N, Fernando WJN, Kim J. Technologies for production of biodiesel focusing on green catalytic techniques: a review. Fuel
Process Technol 2009;90:1502e14.
[10] Chiou B, El-Mashad HM, Avena-Bustillos RJ, Dunn RO, Bechtel PJ, McHugh TH,
et al. Biodiesel from waste salmon oil. Trans ASABE 2008;51:797e802.
[11] Gerpen JV. Biodiesel processing and production. Fuel Process Technol
2005;86:1097e107.
[12] Meher LC, Vidya Sagar D, Naik SN. Technical aspects of biodiesel production
by transesterication e a review. Renew Sust Energ Rev 2006;10:248e68.
[13] UNE-EN 14103. Fat and oil derivatives-Fatty Acid Methyl Esters (FAME)determination of ester and linolenic acid methyl ester contents. European
Committee for Standardization.
[14] Camacho-Paez B, Robles-Medina A, Camacho-Rubio F, Gonzlez-Moreno P,
Molina-Grima E. Production of structured triglycerides rich in n-3 polyunsaturated fatty acids by the acidolysis of cod liver oil and caprylic acid in
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]