PRODUCTION AND OPTIMIZATION OF BIODIESEL FROM MIXED FEED STOCK OIL (KARANJA

AND NEEM) USING RESPONSE SURFACE METHODOLOGY

COLLEGE : SIDDAGANGA INSTITUTE OF TECHNOLOGY, TUMKUR

DEPARTMENT : BIOTECHNOLOGY
GUIDE : MS. NAVYA P N
STUDENTS : SIDDHARTH N R
VINAY KUMAR P
SHWETHA D
SIDDARAJU R
Keywords: Biodiesel, Mixed feedstock oil, Optimization, Response surface methodology

Introduction / background:
Environmental pollution and diminishing supply of fossil fuels are the key factors leading to search for the
alternative sources of energy. Today, 86% of the world energy consumption and almost 100% of the
energy needed in the transportation sector is met by fossil fuels. Since the world’s accessible oil
reservoirs are gradually depleting, it is important to develop suitable long-term strategies based on
utilization of renewable fuel that would gradually substitute the declining fossil fuel production. In addition,
the production and consumption of fossil fuels have caused the environmental damage by increasing the
CO2 concentration in the atmosphere (Westermann et al., 2007). Currently, the most often-used type of
biodiesel fuel is vegetable oil fatty acid methyl esters produced by trans-esterification of high quality
vegetable oil by methanol. Biodiesel derived from vegetable oil and animal fats is being used in USA and
Europe to reduce air pollution and dependency on fossil fuel. In USA and Europe, their surplus edible oils
like soybean oil, sunflower oil and rapeseed oil are being used as feed stock for the production of
biodiesel (Ramadhs et al., 2004 ; Sarin and Sharma, 2007). Since more than 95% of the biodiesel is
synthesized from edible oil, there are many claims that a lot of problems may arise. By converting edible
oils into biodiesel, food resources are actually being converted into automotive fuels. It is believed that
large-scale production of biodiesel from edible oils may bring global imbalance to the food supply and
demand market. Recently, environmentalists have started to debate on the negative impact of biodiesel
production from edible oil (Butler, 2006). There has been significant expansion in the plantation of oil
crops for biodiesel in the past few years in order to fulfil the continuous increasing demand of biodiesel.
Although there is continuous increase in the production of vegetable oil; however, the ending stocks of
vegetable oils are continuously decreasing due to increasing production of biodiesel. Eventually, with the
implementation of biodiesel as a substitute fuel for petroleum-derived diesel oil, this may lead to the
depletion of edible-oil supply worldwide. In order to overcome this devastating phenomenon, researches
have been made to produce biodiesel by using alternative or greener oil resources like non-edible oils.
The non-edible vegetable oils such as Madhuca indica, Jatropha curcas and Pongamia pinnata are found
to be suitable for biodiesel production under the experimental conditions (Meher et al, 2006; Sent et al.,
2003). In this connection, the present study aimed at producing biodiesel from mixed feedstock (karanja
and neem) oil and to optimize the conditions for production of biodiesel using response surface
methodology.

Objectives:
• Production of Biodiesel from mixed feedstock oil (karanja and neem)
• Optimization of process parameters by response surface methodology to increase the yield
• Characterization of biodiesel

Methodology:
Materials: Neem & karanja seeds, sodium hydroxide, sulphuric acid, methanol, potassium hydroxide,
isopropyl alcohol, phenolphthalein indicator
Methods: The oil was extracted from karanja and neem seeds using oil expeller. The oil samples were
determined for their acidity degree (FFA) as per Saad (Saad et.al, 2006). Since the FFA content of the
raw oil was more than 4%, double stage process (Esterification + Trans-esterification) was undertaken.
The esterification reaction was carried out using acid catalyst (sulphuric acid) for reducing the free fatty
1
acid content for production of esters. After the reduction of FFA, trans-esterification reaction was carried
out using methanol and base catalyst (KOH). The optimization step was carried out to enhance the
production of biodiesel. A five level four factor central composite design consisting of 30 experimental
runs including 6 replicates at the center point to allow a better estimate of the experimental error. The
0
design variables were molar ratio (A), catalyst concentration (B, wt %), temperature (C, C) and reaction
time (D, min) while the response variable was biodiesel yield (X, %). The range and the levels of the
independent variables chosen for the current study are presented in Table below:
Independent Variable Range and level
-α -1 0 1 α
Molar Ratio (A) 3:1 4.5:1 6:1 7.5:1 9:1
Catalyst Concentration (wt %) (B) 0.6 0.8 1.0 1.2 1.4
0
Temperature ( C) (C) 40 48 55 62 70
Time of reaction (min) (D) 30 45 60 75 90

Each experiment was performed in duplicates and the average yield of biodiesel was taken as the
response variable X. A second-order polynomial equation was then fitted to the data by a multiple
regression procedure. Statistical analysis of the data was performed by design Package, Design-Expert
version 9.0, from Stat Ease, Inc., Minneapolis, USA to evaluate the analysis of variance (ANOVA), to
determine the significance of each term in the equations fitted and to estimate the goodness of fit in each
case. The fitted polynomial equation was then expressed in the form of three-dimensional response
surface plots to illustrate the main and interactive effects of the independent variables on the dependent
ones. Validation of the model was carried out to check the adequacy of the model. After trans-
esterification step, methanol was recovered using simple distillation set up and was washed using distilled
water. The produced biodiesel was dried and stored for characterization studies. The fuel properties such
as flash point, fire point, kinematic viscosity, cloud point, pour point, calorific value was determined.

Results and Conclusions:

The mixed oil was subjected to free fatty acid estimation and the FFA content was found to be 9%. Since
the FFA content was >4%, esterification step was carried out using acid catalyst and was reduced to
1.6%. Then, the trans-esterification step was carried out according to the experimental runs suggested by
2
Design expert. The regression equation obtained by ANOVA showed that the R value (multiple
correlation coefficient) was 0.9888 (a value > 0.75 indicates fitness of the model). This number is an
estimate of the fraction of overall variation in the data evaluated by the model and the model is thus
capable of explaining 98.88 % of the total variations in the observed response or by experimental
parameters and their interactions. The rest (1.12 %) of the total variation was not explained by the model.
2 2
The adjusted R was 0.9784 and the predicted R was 0.9377, which indicate that the model is good. For
a good statistical model, the R2 should be in the range of 0–1.0 and the nearer to 1.0 the value is, the
more fit the model is deemed to be. The adequate precision value of the model is 43.30. The lower value
of the coefficient of variation (C.V % = 1.74) showed that the experiments were precise and reliable. The
Model F-value of 94.93 implies the model is significant. The p value of the model was < 0.0001 (p < 0.05)
and insignificant lack of fit suggested good fit. According to the p-value (the value, in case of below 0.05,
2
indicated significance level) B, C, D, BC, CD and C were significant. The optimized variables were found
using desirability objective function that assigns relative importance to the responses. Solutions with
higher desirability gave optimum methanol: oil ratio as 4.2, reaction time 45 min, catalyst concentration
o
1.115 wt% and temperature of 62 C for biodiesel production. At these optimized conditions, the maximum
of 89.51% biodiesel was achieved. The viscosity of mixed feedstock biodiesel was 5.9 Cst, flash point
0 0 3
and fire point was 193 C and 205 C and a density of 890 kg/m .
In conclusion, CCD approach with four input variables was used to optimize the biodiesel yield. A yield of
90% of given mixed oil sample (Karanja and Neem) was obtained for a molar ratio of 4.2, a reaction time
0
of 45 minutes, a reaction temperature of 62 C and a catalyst concentration of 1.1 wt.%. Approximately
11% increase in the biodiesel yield was observed using central composite design approach. The fuel
properties of produced biodiesel were closer to the ASTM standard of biodiesel. Therefore, it could be
2
concluded that mixed oil biodiesel could be used as a fuel by blending it with commercial diesel to run
compression ignition engine.

Scope for future work:

Biodiesel and nonedible oils may not eradicate the world’s energy problem, yet it could be a good fuel
additive and alternative fuel for many uses. Biodiesel can be used further for engine performance studies
and various physical properties check like oxygen stability test and sediments minimization techniques.
The performance studies can be extended to multi-cylinder diesel engines, tractor engines and other
diesel engines used in agriculture and transport sector. However, the long term endurance test is also
necessary to evaluate the durability of the engine with prolonged operations. In addition to this, the
improvement in production of mixed feedstock biodiesel should be performed in the future to promote
mixed feedstock biodiesel properties and quality and more research and development in biodiesel
resources and engine designs are needed. Subsequently, further investigations on the additional fuel
property measures and wear analysis of karanja and neem biodiesel fuelled engine is also necessary. In
spite of popular oils like jatropha, pongamia, mahua etc, which have been recommended by the Planning
Commission as a source of biodiesel production, some un-tapped, un-explored vegetable oils like neem,
karanja etc. could also be used to produce biodiesel to fulfill the energy needs of the country as an
alternative or substitute fuel for diesel engines. The small partial replacement of diesel with biodiesel will
alleviate the pressure on existing diesel oil resources and decrease import case of diesel fuel. Moreover,
it is expected that the price of biodiesel will be lower than the price of conventional diesel fuel in the near
future due to the linear increase in the price of conventional diesel fuel with the increase in its demand
and limited supply. The byproduct glycerol can be used for production of value added products, which
could enhance economic viability of biodiesel production. The combination of using lower cost feedstock
with production of value added co-products may greatly enhance biodiesel industry profitability. Research
should focus on developing novel catalytic processes for biodiesel production, such as heterogeneous
catalysis, which can not only be suitable for low quality feedstock with high FFA, but simplify downstream
separation and improve the purity of by-product crude glycerol. Researchers should also incorporate bio
refinery concept, genetic and metabolic engineering knowledge to further study the application of
microalgae as biodiesel feedstock, production of biodiesel using industrial lipases. Positive results will be
probably achieved over the next few years if the research efforts will continue at the same rate as today.