2011 International Conference on Alternative Energy in Developing Countries and Emerging Economies
- 72 -
TABLE VI
R
ESPONSE TABLE OF THE SECOND STAGE EXPERIMENT
Level
Factors
A
B
C
AxC
D
AxD
E
1
1.25 1.03 0.86 0.98 1.13 0.97 1.04
2
0.70 0.92 1.09 0.97 0.82 0.98 0.91
Delta
0.54
0.11
0.23
0.02
0.31
0.01 0.13
Order
1
5
3
6
2
7
4
Select
A2
B2
C1
D2
E2
P
= T + (A2-T) + (D2-T) + (C1-T)
P
= A2 + D2 + C1 - 2T
P
= 0.70+0.82+0.86-2(0.98) = 0.42
Fig. 3. The response graph of the second experiment.
The results of the first stage are conducted following
data analysis and shown that significant order of factors
is catalyst amount, reaction time, reaction temperature,
molar ratio of ethanol: FFA and amplitude of acoustic
power, respectively. According to delta value, amplitude
can be left by considering the one-haft rule of thumb; we
want to consider a logical breaking point as comparing
each factor or interaction with the next strongest effect.
Therefore only four retained factors could be taken to
calculate the
P
-value, but it has a negative value. If we
consider their interactions, there have interactions
between AxC and AxD (the interaction graphs were not
demonstrated at here). So these interactions should be
taken into account, it may be get more corrected results.
Anyhow this experiment can be exposed that catalyst
amount is the most effective factor which can reduce
FFA content through higher level. All effective factors
have the same fashion of reducing FFA content, contrast
with amplitude of acoustic power. Consequently, we had
proceeded the second stage experiment using the first
stage results by re-ordered significant factors, put
interaction into an account and assigned new levels. The
results were shown in a table IV. Data analysis was
performed as same fashion as the previous results. The
order of significant factors is catalyst amount, molar
ratio of ethanol: FFA and reaction temperature,
respectively. The predicted FFA is positive value and the
result of confirmation run is closed to the predicted FFA.
It can be summarized that effect of significant factors
may be changed depending on a level range, but the
strongest factor is not likely changed. The interactions
also depend on a level range. There are different
interactions if the experiments are conducted at different
level range.
V. C
ONCLUSIONS
The results shown that we can do the batch process
esterification of crude palm oil with ethanol assisted
under ultrasonic irradiation with initial FFA content of 5%
until it can be reduced to less than 0.5% wt. The
optimum conditions of this esterification are 60%
catalyst by wt. of FFA, molar ratio of ethanol: FFA of 30,
reaction temperature at 60
°
C, within 2 hours and
amplitude of acoustic power at 75%. Anyhow the
reaction time and amplitude can be reduced because of
insignificant factors. The advantage of this method is no
need outsourcing heat for the reaction. If this reaction is
compared with methanol esterification, it required less
reaction time to achieve a lower FFA content. Finally,
this method is a good alternative way for pretreatment of
fatty acid of triglyceride before transesterification
process.
A
CKNOWLEDGEMENT
The authors would like to thank the National Research
Council of Thailand (NRCT) for funding the project. The
authors also thank Prince of Songkla University, for the
financial support.
R
EFERENCES
[1] Hanh, Hoang Duc, Dong, Nguyen The, Okitsu, Kenji,
Nishimura, Rokuro and Maeda, Yasuaki.,
Biodiesel
production by esterification of oleic acid with short-
chain alcohols under ultrasonic irradiation
condition
, Renewable Energy, vol.34, 2009, pp.
780
–
783.
[2]
Oil yilelds and characteristics
, Available online:
http://
[3] Peace, G. S.,
Taguchi Methods:A Hands-on
Approach
, Addison-Wesley Publishing Co. New
York (USA),
522
p.
[4] Rokhina, Ekaterina V., Lens, Piet and Virkutyte,
Jurate.,
Low-frequency ultrasound in biotechnology:
state of the art
.
Trends in Biotechnology, vol. 27(5),
2009, pp. 298-306.
[5] Srivastava, A. and Prasad, R.,
Triglycerides-based
diesel
fuels, Renewable and Sustainable Energy
Reviews, vol. 4, 2000, pp. 111-133.
[6] Vanichseni, T., Intaravichai, S., Saitthiti, B. and
Kiatiwat, T.,
Potential Biodiesel Production from
Palm Oil for Thailand
, Kasetsart Journal, Vol. 36(1),
2002, pp. 83-87.
0.60
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1
2
Level
FFA (%)
A
B
C
AxC
D
AxD
E
