Evaluation of the effectiveness of the optimization

(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This is an open access article which appeared in a journal published by Elsevier. This article is free for everyone to access, download and read. Any restrictions on use, including any restrictions on further reproduction and distribution, selling or licensing copies, or posting to personal, institutional or third party websites are defined by the user license specified on the article. For more information regarding Elsevier's open access licenses please visit: http://www.elsevier.com/openaccesslicenses

Author's Personal Copy South African Journal of Chemical Engineering 25 (2018) 169–177

Contents lists available at ScienceDirect

South African Journal of Chemical Engineering journal homepage: www.elsevier.com/locate/sajce

Evaluation of the effectiveness of the optimization procedure with methanolysis of waste oil as case study

T

T.F. Adepojua,∗, E.N. Udoetuka, B.E. Olatunbosunb, I.A. Mayena, R. Babalolaa a b

Chemical/Petrochemical Engineering Department, Akwa-Ibom State University, Ikot Akpaden, Mkpat Enin L.G.A, Akwa-Ibom State. Nigeria, P.M.B 1167, Uyo, Nigeria Agricultural Engineering Department, Akwa-Ibom State University, Ikot Akpaden, Mkpat Enin L.G.A, Akwa-Ibom State. Nigeria, P.M.B 1167, Uyo, Nigeria

A R T I C LE I N FO

A B S T R A C T

Keywords: Transesterification Calcinated local white stone Modeling Optimization Statistical analysis Biodiesel

Local white stone (LWS) identified as Brette Pearl Spar Mable was used to catalyze the transesterification of waste used oil (WUO) to waste used oil biodiesel (WUOB) in the presence of methanol acting a solvent. The conversion of WUO to WUOB was monitored by calcinated LWS (CLWS). Analysis of CLWS demonstrated that potassium (K) is the major active component responsible for the activity of the catalyst in CLWS synthesis, and the transesterification of WUO to WUOB in the presence of methanol shows the catalyst to be suitable for biodiesel production from WUO. To model and optimize the process condition, response surface methodology and Artificial Neural Network was employed in the conversion of WUO to WUOB using CLWS as heterogeneous base catalyst. The modeling was carried out by considering three factors, reaction time (RT), catalyst amount (CA) and the ratio of methanol/oil (M/OR). The optimum conditions that achieved 92.45 (%w/w) for RSM and 98.46 (%w/w) for ANN were RT of 40 min, CA of 6 g and M/OR of 5.5:1. Characterization of the produced WUOB shows that the WUOB can replace conventional diesel when blends. Nevertheless, statistical analysis showed that the conversion using presoaked CLWS, proved to be a suitable heterogeneous catalyst for transesterification of biodiesel.

1. Introduction Biodiesel has long been brought to the limelight as a stepping stone to the world's source of renewable energy, it was based on this knowledge that researchers have thrived in their search for optimum conversion of agricultural waste into biodiesel as a renewable energy. According to American Society for Testing and Materials (ASTM), biodiesel is best described as the production of long-chain mono-alkyl esters of fatty acids from renewable lipid feedstock, such as animal fats, waste used oil or vegetable oil, by transesterification reaction (Carlos et al., 2011). It is worthy to know that biodiesel significantly represents approximately 78% of CO2 because it is majorly derived from renewable biomass sources; this fuel diesel produced from renewable lipid feedstock is known for its low emission of pollutants and its biodegradability (Ibrahim and Abubker, 2015). Over the years, waste used oil (WUO) which is one amongst the many lipid feedstocks has posed a lot of challenges to human existent in terms of its disposal. Indiscriminately discarding waste oil has caused environmental problems which severely contaminate the groundwater, plants, and animals, who in turn can possibly suffocate from the depletion of oxygen and wear and tear of sewer pipes (Conrad and R,

∗

2015). It is as a result of these harmful effects, that there is a necessity for the recycling of WUO and maximally converting it to renewable energy (biofuel). The most efficient and economical means of production of biodiesel is the use of lipids feedstock with a high percentage of free fatty acid and the addition of an excess amount of alcohol (Betiku and Adepoju, 2012). Optimization and modeling of the biodiesel produced from WUO will further enhance its yield of the biodiesel (Adepoju and Eyibio, 2016), as well as its optimum conversion. However, research has shown that there is a glitch in the results obtained from using a single variable method for modeling and optimization; hence there is the need to combine two or more methods in modeling and optimization of the process variables needed for biodiesel production (Seramen et al., 2010). Various researchers have used waste oil or waste used oil for biodiesel production and obtained good conversion yields (Chesterfield et al., 2012; Wanodya and Arief, 2013; Kao-Chia et al., 2014; Maurizio et al., 2014; Atilla et al., 2015; Ricky et al., 2015). Since the use of WUO for biodiesel production can conveniently solve the problem of WUO disposal, this paper explores the application of WUO to produce a biodiesel using heterogeneous based catalyst. The catalyst, LWS

Corresponding author. E-mail addresses: [email protected], [email protected] (T.F. Adepoju).

https://doi.org/10.1016/j.sajce.2018.05.002 Received 19 December 2017; Received in revised form 15 May 2018; Accepted 22 May 2018 1026-9185/ © 2018 The Authors. Published by Elsevier B.V. on behalf of Institution of Chemical Engineers. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).


T.F. Adepoju et al.

2.2.4. Base catalyst transesterification of WUO The properties of WUO showed that the acid value of 0.34 mg KOH/ g oil which corresponds to a free fatty acid value (FFA value) of 0.17% was estimated. These observations showed that the oil could be converted directly to biodiesel using catalytic base (known as transesterification). Hence, biodiesel production was carried out using transesterification step as follow. A 500 ml reactor flask was placed on a hot plate with the magnetic stirrer and 200 ml of oil was charged into the reactor for preheating prior to transesterification to remove unwanted moisture content. A mixture of a known weight of base catalyst and a known volume of an alcohol was added to the preheated oil on the hot plate with the magnetic stirrer and the flask was covered with a stopper to prevent alcohol escaping as the reaction proceeded at the desired temperature for complete reaction at a particular time. At the end of the reaction, the resulting mixture was transferred to a separating funnel and allowed to stand for 24 h for glycerol and biodiesel clarity separation. Glycerol was then tapped out of the bottom of the funnel while the untapped biodiesel left in the separating funnel was washed with ionized water to remove the impurities and excess methanol left in the biodiesel. The washed biodiesel was then dried over the heated salt of calcium, allowed to cooled and then filtered to obtained waste used oil biodiesel (WUOB). The final product of the biodiesel yield was then expressed using equation (1).

identified as Brette Pearl Spar Mable was pre-soaked in methanol and calcined before it was used as a biobase for biodiesel production. For modeling and its optimization, an integrated (response surface methodology (RSM) and artificial neural network (ANN)) approach was adopted to determine the effects of variables on the optimum conversion of oil to biodiesel. 2. Materials and methods 2.1. Materials The WUO (groundnut oil and palm oil) used for this research was obtained from Akwa Ibom State University Cafeteria, Nigeria. The obtained WUO was dark and dirty because of it used and re-used over times for frying; this oil was dumped at the back of the cafeteria, used to ignite fire and also littered the ground, hereby polluted the soil because of lack of proper disposal. This oil was collected, pre heated to reduce the viscosity and then filtered to remove the dirt's. The physical, chemical and fuel properties of the WUO were determined using standard AOAC methods. The local white stone (LWS), used as a base catalyst was also collected from the Akwa Ibom University Senate building. The LWS was washed with ionized water, oven dried to a constant weight, crushed and then sieved into a 0.5 cm particle size powder. The powder was stored in a cleaned crucible for further processing.

WUOB % (w / w ) =

Weight of biodiesel produced (g ) Weight of oil sample (g )

× 100

(1)

2.2. Method 2.2.5. Design of experiment for transesterification of WUO to WUOB For the purpose of experimental design, variable factors need to be considered in other to model and optimize the transesterification step. Some of the factors are controllable while others are uncontrollable variables (Adepoju and Olawale, 2015). Factors such as reaction time, catalyst amount, reaction temperature and methanol/oil molar ratio are controllable variables that predict the rate determining step as the reaction proceed. In this study, these factors were selected as reaction time (RT): 40–60 (min): X1; catalyst amount (CA): 4–6 (g): X2; and methanol/oil ratio (M/OR): 3–8 (v/v): X3. Meanwhile, a 23 full factorial design and central composite design will produce 20 experimental runs (Box and Behnken, 1960), hence to reduce the number of experimental runs, Box-Behnken design (BBDRSM) was employed to generate 17 experimental runs used to study the effects of selected factors on biodiesel yield. To ascertain the visibility of using the design, an artificial neural network was also used and genetic algorithms network (GAANN) was tested for the interaction of selected factors on the yield. Table 1 showed the selected factors and their levels. For the coefficient of the quadratic model of the response fitting, multiple regressions model was adopted using Statistical software 10 (Stat Inc., Tulsa, OK, USA). Regression analysis and test of significance are the computational intensive process that is best carried out via statistical software version 10; hence the quality of the fitted model was evaluated using test of significance and regression analysis of variance (ANOVA) via equation (2).

2.2.1. Characterization of waste used oil (WUO) The qualities of the oil determine the nature of reaction (esterification/transesterification or both) to which the process route will take. Among these qualities are the dependable factors that determine the reversibility of the reaction, such factors could be the moisture content, the specific gravity, the acid value and the free fatty acid value. These qualities were determined using recommended AOAC, 2000 standard methods. 2.2.2. Gas chromatography-mass spectrometry analysis of WUO Gas chromatography-mass spectrometer (GCMS), model 19091S433HP-5MS was employed to determine the fatty acid composition of the WUO before transesterification. The analytical conditions of GCMS for fatty acid detection were 30 mm × 250 μm × 0.25 μm, composed of 5% phenyl methyl silox), operating in Electron Multiplier Volts 1329.412 eV, Helium (99.99%) was used as carrier gas at a constant flow of 1.5 ml/min and an injection volume of 1 μl was employed (Split ratio of 10:1), Injector temperature of 150 °C and Ion-source temperature of 250 °C. The oven temperature was programmed from 35 °C (Isothermal for 5 min), with an increase of 4 °C/min, to 150 °C, for 2 min, then 20 °C/min to 250 °C, for 5 min (Isothermal at 250 °C). Mass spectra were taken at an average velocity of 44.297 cm/s with hold up time of 1.1287 min, a pressure of 11.604 psia and frequency of 50 Hz. The overall total GCMS running time was observed for 45 min.

k

2.2.3. Catalyst calcination and elemental characterization Four samples labeled A, B, C, D of LWS grounded powder, each weighs 50 g was measured into 250 ml conical flask, and 100 ml of methanol (BDH Analar: 95%) was added to each flask, shake for 10 min and then filtered. The filtrate was discarded, while the residual cake were calcinated in a Carbolite AAF1100 furnace at 700°C for 4 h for sample A, 5 h for sample B, 6 h for sample C and 7 h for sample D. Furthermore, samples analysis were carried out using a recommended silver standard XRF Spectrophotometer (EDX 3600B) calibrated with ore standard calibration curve. The calcined presoaked powdered CLWS with the highest potassium (K) was used for base catalysed transesterification process.

RF = τ0 +

k

k

∑ τi Xi + ∑ τii Xi2 + ∑ τij Xi Xj + e i=1

i=1

(2)

i 0.05, thus, make it insignificant model term. The significant of the variable factor is fit for the adequate representation of the variables considered in the experimental design ( X1,X2 and X3 ) at a confidence level of 95%. Furthermore, the pvalue of X2 with an F-value of 12.70 shows that X2 is the least significant variable when compared with the other variables (sources). The analysis of variance of regression equation model is as shown in Table 5 (ANOVA). The root sum of the square (SSR), otherwise known as the model is highly significant with p-value < 0.0001 and Fvalue = 373.29. The residual error sum of the square (SSE) and the total sum of the square (SST) produced no value for p and F-values,

df

Mean Square

F-value

p-value

X1 X2 X3 X1X2 X1X3 X2X3 X12 X22 X32

51.00 7.61 126.41 434.72 0.12 531.30 357.74 287.80 235.74

1 1 1 1 1 1 1 1 1

51.00 7.61 126.41 434.72 0.12 531.30 357.74 287.80 235.74

85.17 12.70 211.08 725.92 0.20 887.19 597.36 480.58 393.65

< 0.0001 0.0092 < 0.0001 < 0.0001 0.6648 < 0.0001 < 0.0001 < 0.0001 < 0.0001

df

Mean Square

F-value

p-value

SSR (model) Residual Lack of Fit SSE (pure error) SST (cor total)

2011.91 4.19 2.18 2.01 2016.10

9 7 3 4 16

223.55 0.60 0.73 0.50 –

373.29 – 1.44 – –

< 0.0001 – 0.3550 –

WUOBY (% w / w ) = 68.24 − 2.52X1 + 0.98X2 − 3.98X3 + 9.22X12 + 8.27X22 − 7.48X32 − 10.43X1 X2 + 0.17X1 X3 − 11.53X2 X3

(7)

To test the fit of the model equation, the regression model was determined by computing R2, R 2 , RSME and ADD of the model (Table 7). R2 which provides a measure of how much variability in the WUOBY values can be explained by the RT, CA and M/OR and their interactions (Sudamalla et al., 2012). The value of R2 value is always Table 6 Regression coefficients and significance of response surface quadratic.

Table 4 Test of significance for every regression coefficient. Sum of squares

Sum of squares

while the lack of fit with p-value = 0.3550 is not significant due to low F-value of 1.44. Insignificant lack of fit is better for the fitness of the model. Shown in Table 6 is the regression coefficients and significant of response surface quadratic value. The significance of regression was evaluated by F-value and p-values using Fischer's and null-hypothesis tests. The F-value predicts the quality of the entire model considering all design factors at a time, whereas the p-value is the probability of the factors having very little or no effects on the experimental data (Panwal et al., 2011). Based on the report, the regression model of Fvalue = 2011.91 is highly significant with p-value < 0.0001, respectively. The lack of fit value observed in the table showed insignificant effects on the response (WUOBY). Larger F-value signifies better fit of the model to the experimental data (Panwal et al., 2011). The regression model of F-value = 2011.91 with p-value < 0.0001, obtained in this analysis shows that the statistical software used for the modeling and optimization is good and reliable. The mathematical relationship among the selected variables and the observed yield is expressed was previously elucidated in Eqn. (4). This mathematical relationship among the selected variables and the observed yield is expressed in Eqn. (7).

3.4. Synthesis of WUOB

Source

Source

Factor

Coefficient Estimate

df

Standard Error

95%CI Low

95%CI High

VIF

Intercept X1 X2 X3 X1X2 X1X3 X2X3 X12 X22 X32

68.24 −2.52 0.98 −3.98 −10.43 0.17 −11.53 9.22 8.27 −7.48

1 1 1 1 1 1 1 1 1 1

0.35 0.27 0.27 0.27 0.39 0.39 0.39 0.38 0.38 0.38

67.42 −3.17 0.33 −4.62 −11.34 −0.74 −12.44 8.33 7.38 −8.37

67.42 −3.17 0.33 −4.62 −11.34 −0.74 −12.44 8.33 7.38 −8.37

– 1.00 1.00 1.00 1.00 1.00 1.00 1.01 1.01 1.02

Table 7 Fitness of the model.

173

Source

R2

R2

RMSE

ADD

Predicted WUOB (% w/w)

BBDRSM GAANN

0.9979 0.9950

0.9967 0.9999

0.7089 0.6971

1.06 2.7728

92.45 98.46


T.F. Adepoju et al.

Fig. 3. (a–c): 3-dimensional plots of WUOBY showing by BBDRSM and GAANN.

174


T.F. Adepoju et al.

Fig. 4. Experimental and the predicted value by BBDRSM and GAANN.

The graphical representation of the plots of the observed yield and the predicted value by BBDRSM and GAANN are shown in Fig. 4. The perfectly fixed points on the graphs showed that the residual values are so small and do not affect the good fit of the model data.

between 0 and 1 (Haider and Pakshirajan, 2007; Liu and Wang, 2007). The value of 0.9979 obtained for BBDRSM and 0.9950 for GAANN is very high indicated a high R2 figure for a good model. However, R2 cannot verify whether the coefficient ballpark figure and its prediction are prejudiced. It also does not show an R2 figure for a good model, or a high R2 figure for a model that doesn't fit. However, R 2 compensates for the addition of variable and only increases if the new term enhances the model above what would be obtained by probability, and decreases when the predictor enhances the model less than what it predicted by chance. Hence, R 2 is the best estimate of the degree of relationship among the selected variables. The R 2 of 0.9999 obtained by GAANN indicated the superiority of GAANN over BBDRSM in terms of fitness of model. The root mean square error (RSME) of GAANN is lesser than that of BBDRSM, but the ADD which accounted for overfitting condition, of GAANN is higher than that of BBDRSM. The higher the value, the better the fitness, this also supports the earlier statement that GAANN performs more better than BBDRSM, but both statistical analysis software's are good for experimental modeling and optimization when require.

3.5. Quality characterization of WUOB The qualities of biodiesel (WUOB) obtained from transesterification of WUO using CLWS was obtained by using standard test methods, and were compared with other results obtained for biodiesel qualities from different oil sources as well as compared with ASTM D6751 and EN 14214 are shown in Table 8. At 28 °C, WUOB was light-brown liquid in colour with the moisture content of 0.014%. The specific gravity and refractive index of the WUOB were 0.85 and 1.4985, respectively. Observation on specific gravity showed that the value agreed with what was reported by other researchers for the same process but different oil source (Table 8). Meanwhile, the percentage free fatty acid obtained was 0.17 corresponds to the acid value of 0.34 mg of KOH/g. These values correspond to the biodiesel standard prescribed by ASTM D6751 and EN 14214 (acid ≤ 0.50 mg of KOH/g). Iodine value of WUOB of 55.40 g I2/100 g of sample is far below the maximum limit of 120 prescribed in EN 14214, but was closer to what was reported by (Deka and Basumatary, 2011). The saponification value, which showed the nature of the fatty acids constituent of WUOB and thus, depends on the average molecular weight of the fatty acids constituent of the oil, was obtained as 104.7 mg KOH/g sample. The value indicated the nature of the WUOB produced. The average molecular weight computed in this study was in line with was reported by (Kathleen et al., 2014). The peroxide value of 1.50 (meq O2/kg sample) show that the WUOB is clean and free from oxidative rancidity at room temperature accounted for it stability (Adepoju and Eyibio, 2016). Other additional fuel properties, such as cetane number, which measures the fuel's ignition delay and combustion quality, the higher the cetane number, the shorter the delay interval and the greater the combustibility. Fuels with low cetane number are difficult to start, hence it smokes. Standard minimum specification value of cetane number for biodiesel is within the range of 47–51 (ASTM D6751 and EN 14214). The WUOB cetane number obtained in this study was 60.50 which show the WUOB to be of high fuel potential and of great combustibility. The diesel index of

3.4.2. Variables effects on WUOBY The variation or the interaction of the factors on the WUOBY was investigated by plotting the three dimensional graphs (Fig. 3) using Design Expert 10 software (StatSoft, Inc. Tulsa, Ok. U.S.A) and NeuralPower_21356. The curvature nature of the plotted graphs indicated significant interactions among the considered factors. Fig. 3a shows the interactions of CA and RT on the WUOBY. It was observed that the yield increases as the CA increases and RT decreases. Unlike Fig. 3b, the plot showed the minimum WUOBY value at the RT range of −1 to −0.5 and M/OR ranges 0.5 to 1, -1 to −0.6. The lower M/OR of −1 and higher CA of 1 favoured the high WUOBY (Fig. 3c), but the value ranges lower WUOBY showed in Fig. 3a. The variable factors examined in this study have been confirmed by researchers to have greater effects on the yield of biodiesel (Betiku et al., 2015; Sunil and Sangeeta, 2015; Hameed et al., 2009; Rajendra et al., 2009; Vega et al., 2014). For a reversible reaction such as the production of biodiesel, which is subjected to equilibrium control of reaction, the methanol oil ratio must be greater than the stoichiometric amount to have a good conversion of the biodiesel (Betiku et al., 2016). The range (5.5–8) provided by this work proved adequate for the optimum conversion of WUO to WUOB. 175


T.F. Adepoju et al.

Table 8 Comparative of WUOB qualities with other biodiesel production from different sources. Properties

(Balusamy and Marappan, 2009)

(Dhoot et al., 2011)

(Adebowale et al., 2012)

(Yarkasuwa et al., 2013)

(Betiku and Ajala, 2014)

(American Society for Testing and Materials (ASTM))

(EN 14214, 1421)

WUOB

AV (mg KOH/g oil) FFA (mg KOH/g) IV (g I2/100 g) SV (mg KOH/g) CN MC (wt. %) SG MMM HHV (MJ/kg) DI

– – – – 47 – – – – –

0.057 – 69.9 – 61.5 – – – – –

0.3 – – – – – – – – –

0.2 – – – 54.2 – – – – –

0.46 – – – 123.25 – – – – –

< 0.80 < 0.40 – – 48–65 < 0.03 0.86–0.90 – – 50.4

0.5 max 0.25 max. 120 max. – > 51 0.02 0.85 – – –

0.34 0.17 55.40 104.70 60.50 0.014 0.85 273.58 41.87 74.03

References

74.03 and mean molecular mass of 273.58 obtained were higher than that of neat diesel, indicated low emission of CO, NOX and H2S. The higher heating value, also known as gross calorific value or gross energy of a fuel was obtained as 41.87 MJ/kg. This value takes into account the latent heat of vaporization of water in the combustion products. Hence, the produced WUOB can be classify as a good energy carrier.

Adebowale, K.O., Adewuyi, A., Ajulo, K.D., 2012. Examination of fuel properties of the methyl esters of Thevetia peruviana seed oil. IJGE 9, 297–307. Adepoju, T.F., Eyibio, U.P., 2016. Study on oil extraction from citurllus lanatus oil seeds and its statistical analysis: a case of response surface methodology and artificial neural network. CRJ 1, 28–36. Adepoju, T.F., Olawale, O., 2015. Optimization and predictive capability of RSM using controllable variables in Azadiracha indica oilseeds extraction process. IJCMR 3, 1–10. American Society for Testing and Materials (ASTM). Aramide, M.A., Sherriff, O.A., Eriola, B., 2015. Optimization of Bauhinia monandra seed oil extraction via artificial neural network network and response surface methodology: a potential biofuel candidate. Ind. Crop. Prod. 67, 387–394. Atilla, B., Mert, G., Ihsan, K., Elif, N., Abdulvahap, C., 2015. Determination of transesterification reaction parameters giving the lowest viscosity waste cooking oil biodiesel. Energ. Procedia 195, 2492–2500. Balusamy, T., Marappan, R., 2009. A comparative analysis on the performance and emission characteristics of thevetia peruviana seed oil (TPSO) with other non-edible oil in CI engine. IEJ 10, 93–100. Betiku, E., Adepoju, T.F., 2012. Methanolysis optimization of sesame (Sesamum indicum) oil to biodiesel and fuel quality characterization. Int. J. Electr. Electron. Eng. 4, 2251–6832. Betiku, E., Ajala, S.O., 2014. Modeling and optimization of Thevetia peruviana (yellow oleander) oil biodiesel synthesis via Musa paradisiacal (plantain) peels as heterogeneous catalyst: a case of artificial neural network vs. response surface methodology. Ind. Crop. Prod. 53, 314–322. Betiku, E., Samuel, S.O., Sheriff, O.A., Olatunde, S.O., 2015. Performance evaluation of artificial neural network coupled with genetic algorithm and response surface methodology in modeling and optimization of biodiesel production process parameters from shea tree (Vitellaria paradoxa) nut butter. J. Renew. Sustain. Energy 76, 408–417. Betiku, E., Aramide, M.A., Ojumu, T.O., 2016. Banana peels as a biobase catalyst for fatty acid methyl easters production using Napoleon's plume (Bauhinia monandra) seed oil: a process parameters optimization study. At. Energ. 403, 793–806. Betiku, E., Anieti, O.E., Omoniyi, P., Ojumu, T.O., T. O, 2017. Two step conversion of Neem (Azadirachta indica) seed oil into fatty methyl esters using a heterogeneous biomass based catalysts: an example of cocoa pod husk. Energy Fuels 31, 6182–6193. Box, G.E., Behnken, D.W., 1960. Some new three level design for the study of quantitative variables. Technometrics 2, 455–475. Canakci, M., Van Gerpen, J., 1999. Biodiesel Production via Acid Catalysis. Trans ASAE. Carlos, A., Guerrero, F., Andres, G.R., Fabio, E.S., 2011. Biodiesel Production from Waste Used Oil, Biodiesel, fourth ed. Feedstocks and Processing Technologies. Chesterfield, D.M., Peter, L.R., Essam, O.A., Adesoji, A.A., 2012. Production of biodiesel via ethanolysis of waste cooking oil using immobilized lipase. Chem. Eng. J. http:// dx.doi.org/10.1016/j.cej.2012.07.039. Conrad, R, R., 2015. Used Oil Wreaks Havoc on Sewers. Triple Pundit. Deka, D.C., Basumatary, S., 2011. High quality biodiesel from yellow oleander (Thevetia peruviana) seed oil. Biomass Bioenergy 35, 1797–1803. Dhoot, S.B., Jaju, D.R., Deshmukh, S.A., Panchal, B.M., Sharma, M.R., 2011. Extraction of Thevetia peruviana seed oil and optimization of biodiesel production using alkalicatalyzed methanolysis. J. Altern. Energy Sources Technol. 2, 8–16. EN 14214. Haider, M.A., Pakshirajan, K.K., 2007. Screening and optimization of media constituents for enhancing lipolytic by a soil microoraganism using statically designed experiments. Appl. Biochem. Biotechnol. 141, 377–390. Hameed, B., Lai, L., Chin, L., 2009. Production of biodiesel from palm oil (Elaies guineensis) using heterogeneous catalyst: an optimization process. Fuel Process. Technol. 90, 606–610. Ibrahim, H.G., Abubker, A., 2015. Alshuiref, a. Ahmed, maraie. Recycling of waste cooking oils (WCO) to biodiesel production. JMES 2, 721–725. Kao-Chia, H., Ching-Lung, C., Ping-Xuan, H., Meng-Shan, W., Chien-Chang, H., JO-Shu, C., 2014. Biodiesel production from waste cooking oil by two-step catalytic conversion. Energ Procedia 61, 1302–1305.

4. Conclusion This study evaluates the effects of variables such as reaction time, catalyst amount and methanol/oil ratio on transesterification of WUO to WUOB using calcinated local white stone. The production of biodiesel was optimally produced using combined statistical software, Box Behnken Design and genetic algorithms from response surface methodology and artificial neural network. The highest biodiesel yield of 99.80 (% w/w) was obtained at:

• 40 min reaction time • 6 g of Catalyst amount • 5.5 methanol/oil ratio The predicted values under these same conditions by BBDRSM and GAANN were found to be 92.65 and 99.80 (% w/w), respectively. The experimental validation in triplicate produced 92.45 (% w/w) for BBDRSM and 98.46 (% w/w) for GAANN. The result of the calcination showed potassium (K) to be the major active component responsible for the activity of the catalyst in CLWS synthesis. The gas chromatograph analysis of WUO showed the oil to be highly unsaturated. Funding This research did not receive any specific grant from funding agencies in public, commercial, or not-for-profit sectors. Competing interest The research article is original work of authors and was truly carried out by authors. Authors declare no competing interest whatsoever. Acknowledgements Authors show their profound gratitude to the Staff of Chemical/ Petrochemical Engineering, Akwa Ibom State University for their support and encouragement during this research work. Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx. doi.org/10.1016/j.sajce.2018.05.002. 176


T.F. Adepoju et al.

Environ. Sci. 6 (4), 1178–1185. Seramen, S., Rajendran, A., Thangavelu, V., 2010. Statistical optimization of anticholesterolemic drug lovastatin production by red mold Monascuspurpureus. Food Bioprod. Process 88, 266–276. Sudamalla, P., Saravanan, P., Matheswaran, M., 2012. Optimisation of operating parameters using response surface methodology for adsorption of crystal violet by activated carbon prepared from mango kernel. SERKET 22, 1–7. Sunil, K., Sangeeta, N., 2015. Transformation of waste cooking oil into C-18 fatty acids using a novel lipase produced by Penicillium chrysogenum through solid state fermentation. J. Biotechnol. 5, 847–851. http://dx.doi.org/10.1007/s13205-0140268-z. Vega, P.M., Luna, A.S., de-Figueuedo, P.M., de-Oliveira, V.C., Henrique Zn, C.A., 2014. Al catalyst for heterogeneous biodiesel production: basicity and process optimization. At. Energ. 75, 453–462. Wanodya, A.K., Arief, B., 2013. Synthesis of biodiesel from second used cooking oil. Energ Procedia 32, 190–199. Yarkasuwa, C.I., Wilson, D., Michael, E., 2013. Production of biodiesel from yellow oleander (Thevetia peruviana) oil and its biodegradability. JKCS 57, 377–381.

Kathleen, F.H., Goran, T.V., James, C.R., Zoltan, Nagy, B.S., 2014. Kinetics of the pretreatment of used cooking oil using Novozyme 435 for biodiesel production. Chem. Eng. Res. Des. 92, 713–719. Liu, G.C., Wang, X.L., 2007. Optimization of critical medium components using response surface methodology for biomass and extracellular polysaccharide production by Agaricus blazei. Appl. Microbiol. Tech. 74, 78–83. Maurizio, C., Sonia, C., Silvia, C., 2014. A pilot Scale Study of waste vegetable oil transesterification with alkaline and acidic catalyst. Energ. Procedia 45, 198–206. Panwal, J.H., Viruthagiri, T., Baskar, G., 2011. Statistical modeling and optimization of enzymatic milk fat splitting by soybean lecithin using response surface methodology. IJNM 3, 50–57. Rajendra, M., Jena, P.C., Raheman, H., 2009. Prediction of optimized pretreatment process parameters for biodiesel production using ANN and GA. Fuel 88, 868–875. Ricky, P., Teng-Chien, C., Ming-Chin, L., Aharon, G., Jiunn-Der, L., Yao-Hui, H., 2015. Novel Technology for biodiesel production from cooking and waste cooking oil by microwave irradiation. Energ. Procedia 75, 84–91. Ridha, B., Snoussi, Y., Mounir, B., Manef, A., 2015. Waste frying oil with high levels of free fatty acids as one of the prominent sources of biodiesel production. J. Mater.

177