Using a wire coil insert for biodiesel production enhancement in a
microreactor
Babak Aghel a ,Masoud Rahimi a ,b ,⇑,Arash Sepahvand b ,Mohammad Alitabar b ,Hamid Reza Ghasempour b
a CFD Research Center,Chemical Engineering Department,Razi University,Kermanshah,Iran
b
Department of Biotechnology-Chemical Engineering,Science and Research Branch,Islamic Azad University,Kermanshah,Iran
a r t i c l e i n f o Article history:
Received 16January 2014Accepted 1May 2014
Available online 20May 2014Keywords:Microreactor Biodiesel Wire coil
Process intensification
Response surface methodology
a b s t r a c t
In the present work,the application of wire coil to promote mixing in a microreactor during continuous production of biodiesel was studied.For this aim,soybean oil as a feedstock and potassium hydroxide as a homogeneous catalyzed were used.The influences of the various parameters such as geometric
and oper-ational conditions on the performance of biodiesel production were experimentally examined.Response surface methodology (RSM)in conjunction with the Box–Behnken method was used to statistically ana-lyze and optimize the biodiesel production process.The comparison between two types of reactors (with and without wire coil)shows a significant enhancement in mixing during transesterification.The impacts of different wire coil lengths and wire coil pitchs on methyl ester conversion were also investigated.A reaction yield of 99%at the residence time of 180s was obtained in the modified microreactor.However,the measured pressure drop show that the microreactor equipped with wire coil consumed more energy.Therefore,performance ratio was defined to evaluate energy efficiency and the results show the advan-tage of using the wire coil insert in lower feed flow rates.
Ó2014Elsevier Ltd.All rights reserved.
1.Introduction
In recent years,as a result of the increasing demand for fossil fuels and rapid growth in population studies have been gradually more directed toward looking for renewable fuels.Biodiesel is an important alternative fuel which produced from renewable sources,and it is biodegradable and non-toxi
c.It has much better characterization compared to fossil diesel fuel with no difference in heat of combustion such as cetane number,flash point,and lubricity characteristics [1–5].
In general,fatty acid methyl esters known as biodiesel,derived from vegetable oil or animal fat through the transesterification process.According to the reaction,oil or fat was reacted with alco-hol,such as methanol or ethanol,in the presence of a catalyst to form fatty esters and glycerol [6].However,the high cost of feed-stock such as vegetable oil or animal fats makes biodiesel still more expensive and less economically competitive.Therefore,reducing operational cost of biodiesel process from wastes or renewable materials is more vital.Many topics were highlighted by authors such as the cost of fuel,methods of product,the characterizations of reactor and the effects of operation variable [7,8].
In general,due to availability,economic and fast reaction rate at low temperature sodium hydroxide (NaOH)or potassium hydrox-ide (KOH)are commonly used as an alkaline catalyst to produce methyl ester in transesterification reaction [9,10].However,limita-tion of use oils containing less than 0.5wt.%of free fatty acid (FFA)or with acid values of less than 1mg KOH/g is the main disadvan-tages of these catalysts [11,12].
In conventional methods of Biodiesel production,the reaction was taken places in batch reactors [13],w
hich are usually costly and time consuming.Different technologies have been conducted to promote the efficiency of the biodiesel synthesis and lower the related processing costs [14].Among these methods,some researchers proposed ultrasonic irradiation technique for biodiesel production in batch rectors to obtain high yield with corresponded less residence time [15,16].Furthermore,in another approach,some authors employed microwave to enhance oil transesterifica-tion [17,18].In recent years,supercritical transesterification for high conversion efficiency of biodiesel with have been reviewed [19,20].
The main part of biodiesel process,which increased cost and time is mixing.It appears that the challenge on reducing the mass transfer between the oil and alcohol to increase the rate of transe-sterification is most important.To achieve excellent mixing,pow-erful agitators in a stirred tank to create shear (for mixing)was needed.However,in practical terms,high mechanical stirring
/10.an.2014.05.0090196-8904/Ó2014Elsevier Ltd.All rights reserved.
⇑Corresponding author at:Chemical Engineering Department,Razi University,Taghe Bostan,Kermanshah,Iran.Tel.:+988314274530;fax:+988314274542.
E-mail addresses:masoudrahimi@yahoo ,m.rahimi@razi.ac.ir (M.Rahimi).
power was needed to meet the required reduction in mass transfer. Several studies have recently been published on the effect of mix-ing such as different agitator speeds and type of agitator during transesterification reaction[21–23].
In order to obtain high yield of FAME,static mixer were used to intensify biodiesel process by some authors[24,25].The role of sta-tic mixer as a mixing device is to increase the interaction between the phases and enhances the mass transfer.This was investigated by Somnuk et al.[26,27]and according to their experimental results using static mixer has strong impact on the yield of methyl ester production.
Microreactors,as a new type of reactors,have some attractive characteristics such as high surface-to volume ratio,short diffusion distance,very efficient heat and mass transfer along with intrinsic safety[28].Microreactors can be used in a continuous mode and it is expected to be replaced by many batch chemical processes.In recent years,few studies conducted by researchers on continue production of biodiesel in T or Y structures of microreactor[29–31].Moreover,there are few efforts on modified geometric param-eters of micro-structured mixers for biodiesel synthesis[32–34]. On the other hand,optimization of the reaction parameters was studied in some research[35–37].
In the present work,a microreactor equipped with a wire coil insert has been developed in order to improve the efficiency of mixing in plain one,which was studied previously[38].The effects of main parameters on the performances of produced Fatty Acid Methyl Ester(FAME)such as molar ratio of methanol to oil,tem-perature of reaction,concentration of KOH was experimentally investigated by Box–Behnken method.The response surface meth-odology(RSM)was carried out for optimization of this process.In addition,the effects of various passive micromixers(different kind of wire coil)on the performance of biodiesel synthesis have been investigated.
2.Materials and methodology
2.1.Materials
The soybean oil used in this study was supplied from Nazgol oil (Kermanshah,Iran)and has the average molecular weight,specific gravity and saponification index of863.47g/mol,910kg/m3and 191.88mg of KOH/g oil,respectively.Methanol(purity>99.5%), potassium hydroxide(purities>85%,pellets),sulfuric acid(AR grade)and normal heptane(purity>99.0%)were supplied from Merck hyl laurate(methyl dodecanoate,99.7%)as stan-dard for GC analysis was supplied by Sigma–Aldrich.
2.2.Reactor setup and test procedure
The microreactor consists of two parts including a microtube and a micromixer used in the previous study done by authors [38].The microtube was a stainless steel tube with an outer diam-eter of1.58mm and an internal diameter of0.9mm.A wire coil was embedded in the microtube,as an augmentation device(a sil-ver wire coil with0.8mm average diameter)as shown in Fig.1.
Three different kinds of wire coil have been used to promote mixing as follows:(1)a wire coil with a length of30cm and aver-age pitch length of0.5mm,(2)wire coil with a length of10cm length and average pitch length of0.5mm and(3)wire coil with a length of10cm length and average pitch length of1mm.The employed wire coils inserted into the studied microreactor is shown in Fig.2.
During the experiments,methanol solution and soybean oil were fed to the micromixer.The methanol solution was prepared from methanol and KOH(0.6–1.8wt.%with respect to the oil).In addition,in order to adjust the temperature of reaction,the mic-roreactor was placed in a water bath.
The peristaltic pump(QISÓDSP100)was used to divert the soy-bean oil and methanol solution from different side of micromixer. The pressure drop during experiment was measured using pressure transducers mounted at the inlet and the end of the microreactor.
At T-shaped junction of micromixer,the soybean oil and meth-anol solution mixed to proceed the transesterification reaction. Afterwards,the mixture of solution exit at the outlet of micromixer andfluidflows through the microtube along wire coils.Finally,the mixture from the outlet of microtube was collected in aflask. Based on the characteristic of transesterification reaction,the col-lectingflask was placed in an ice-water bath and the product was neutralized by the acid.Consequently,the collected mixture was poured into a separating funnel and allowed to separate for 20min.The upper layer was collected as main product and the lower layer drained as byproduct and the excess of KOH and meth-anol.In order to assure the accuracy of measurement,the main product that contains the methyl ester,washed three times with hot distilled water to remove excess materials.The moisture of the washed esters was subsequently removed by heating in an oven.
According to chemical stoichiometric ratio,the overall transe-sterification requires3mol of alcohol react with1mol oil to yield 3mol of ester and1mol of glycerol.In practice,since this reaction is reversible,in order to force the reaction toward termination,it is necessary to keep the molar ratio at higher value than the stoichi-ometric value.
The experimental test was performed at three different ratios of methanol to oil(6:1,9:1and12:1),KOH concentrations(0.6%,1.2% and1.8%(w/w))and temperatures(55,60and65°C).In addition, the experimen
ts were usually repeated2–3times to determine the experimental error and the reproducibility of the data.
The biodiesel purity was analyzed using a gas chromatography (GC)to determine the content of the fatty acid esters(FAME).The GC(Varian CP3800Netherland)was equipped with a fused silica capillary column(DB-WAX,30m_250lm_0.25lm,nominal)and aflame ionization detector(FID).For this aim,methyl laoranoate served as the internal standard for GC.
The analysis of FAME%for samples was done by method of Wang et al.[39]according to the following equation:
FAME%¼
area of all FAME
area of all reference
Â
weight of reference
weight of bioiesel sample
Â100
ð1Þ
Nomenclature
D hydraulic diameter,m
L length,m
M concentration,mol/l
D P pressure drop difference,Pa u velocity,m sÀ1T temperature,K t time,s
q density,kg mÀ3
542 B.Aghel et al./Energy Conversion and Management84(2014)541–549
2.3.Experimental design and statistical analysis
Response surface methodology(RSM)is a mathematical tech-nique which has many applications in the optimization and pro-cess design.This method is quite useful for evaluating the effects of independent variables on the response.By using the response surface methodology,the optimum operating condition can be determined[40].
In this study,the transesterification process of soybean oil was performed using Box–Behnken design for a higher yield of biodie-sel.Three independent variables for transesterification reaction are molar ratio methanol to oil(6–12),reaction temperature(55–65°C)and KOH concentration(0.6–1.8wt.%).The ranges and levels of the three independent variables with coded levels of each parameter are listed in Table1.The percent of FAME obtained from the reaction was chosen as a response.In addition,RSM was uti-lized to examine the influence of three variables on the percentage of methyl ester.
A quadratic regression model wasfitted to the
to relate the predicted response variable(FAME%)and
dent variable.The form of the mathematical model
Eq.(2):
Y yield¼b0þ
X3
i¼1
b i X iþ
X3
i¼1
b ii X2
i
þ
X2
i¼1
X3
j¼iþ1
b ij X ijð2Þ
where Y yield is the FAME%,X i and X ij are the uncoded independent variables,b0is the offset term and b i,b ii,b ij are regression coefficients.
Analysis of variance(ANOVA)is a reliable method to analyze and define the degree of certainty of experimental data[41].Fur-ther,the quality of the model was evaluated using the coefficients of determination(R2)and analysis of variance.
reaction massThe complete experimental design matrix corresponding to the Box–Behnken design in terms of independent variable and actual results is presented in Table2.It can be noticed that in order to minimize errors from the systematic trends the experiments were run at random.
3.Results and discussion
The mean values of the percentage of FAME for the correspond-ing factors at each level were calculated according to the
Fig.1.Schematic view of microreactor system.
Fig.2.The wire coil with pitch length of1mm.Table1
Experimental range and levels of variable.
Variables Levels
À10
Temperature(°C)5560 Catalyst concentration(wt.%)0.6 1.2 Molar ratio(methanol/oil)6:19:1
assignment of the experiment.The mean values of the three levels of each factor show how the percentage of FAME changes when the level of that factor is changed.
Moreover,to verify other variables such as length of wire coil and pitch wire coil tests were performed.
3.1.Effect of the molar ratio of oil/methanol
As it is well known,the methanol molar ratio to oil is one of the most important factors that can affect the yield of methyl esters [42].Fig.3presents the effect of molar ratio of methanol/oil on the percentage
of methyl esters of the two different microreactors using three molar ratio levels of6:1,9:1and12:1.As can be seen in thisfigure,both microreactors showed the same tendency. Although the results from Fig.3indicate that the percentage of methyl esters of microreactor with wire coil was higher than plain microreactor at the same methanol to oil ratio.
Results given in thisfigure clearly indicate that molar ratio of oil/methanol is a dominant variable of the process and the percent-age of methyl esters increased with molar ratio from6:1to9:1.It could be assumed that,the reaction was incomplete for a molar ratio of methanol to oil3:1.In other words,excessive molar ratio could speed up the reaction and could ensure complete reaction. The microreactors reached to the highest percentage of methyl esters when molar ratio of methanol to oil was9:1.It can be noticed that increased in molar ratio from6:1to9:1caused4.6% and7.2%improvements in the percentage of methyl ester for mic-roreactor with and without wire coil,respectively.This demon-strates that using wire coil,which enhances mixing,can decrease the effect of molar ratio of methanol/oil on the percentage of methyl ester.On the other hand,an increase in the molar ratio level above9:1led to an unexpected slight reduction in the per-centage of methyl esters.This unexpected result can probably be related to reversible reaction,which shifts the equilibrium limita-tion toward the products.
3.2.Effect of catalyst concentration
The concentration of catalyst is another important factor affect-ing the percentage of FAME.Fig.4shows the effect of catalyst con-centration on the percentage of methyl esters by applying three different levels of catalyst concentration including:0.6,1.2and 1.8wt.%when the residence time was kept at26s.The results indi-cate that increase in catalyst concentration from0.6to1.2wt.%can increase the percentage of produced methyl esters in the plain and wire coil inserted microreactors.As can be seen in thisfigure,at catalyst loading of0.6wt.%the percentage of FAME for the mic-roreactor with wire coil is92.24%whereas for microreactor with-out wire coil is86.95%.The highest percentage of94.5%and 88.7%were obtained at1.2wt.%KOH during26s of residence time for microreactor with and without wire coil,respectively.The rea-son for these differences might be an insufficient mixing of catalyst with reactants in the plain microreactor compared with the plain one.
It should be also noted that when the catalyst concentration increased from1.2to1.8wt.%the percentage of methyl esters did not increase and decreased slightly.This reduction of the yield in the biodiesel production can be observed in the both microreactors.
It can be concluded that due to high content of FFA,the sapon-ification reaction becomes remarkable a
t higher concentration of KOH,and a gel was produced and some product may be encased into the gel.This phenomenon is clearly observed for both microre-actors at same condition.
Table2
Experimental design and results.
Experiment No.Temperature
(°C)
Catalyst
concentration
(wt.%)
Molar ratio
(methanol/oil)
FAME
(%)
165 1.21292.51
2600.6691.86
360 1.8687.92
455 1.2690.17
555 1.21293.33
660 1.2999.95
7600.61291.54
8550.6993.44
960 1.81290.67
1065 1.8989.02
1155 1.8988.72
12650.6992.12
1365 1.2689.87
1460 1.2996.93
1560 1.2998.78
544 B.Aghel et al./Energy Conversion and Management84(2014)541–549
3.3.Effect of reaction temperature
Fig.5shows the effect of temperature on the percentage of FAME of soybean oil in microreactors(with and without wire coil) at three temperature levels of55,60and65°C.As expected, increase in reaction temperature during transesterification reac-tion has a favorable effect on yield of FAME.As it will be seen, the effect of temperature on the percentage of FAME in both mic-roreactors is similar.This might
be due to increase in reaction rate and improve in mass transfer rate by increase in temperature.A moderate increase in temperature over the range of55–60°C can lead to an increase in the percentage of FAME after26s of resi-dence time.However,as shown in Fig.5,at60°C,maximum yields of93.95%and88.67%were obtained for the microreactors with and without wire coil,respectively.It must be pointed out that the reaction temperature must be lower than the boiling point of the alcohol.
It can see that due to high content of FFA in oil,the biodiesel conversion decreased rapidly as the temperature of reaction rises from60°C to65°C.However,the magnitude of reduction does not same for microreactor with and without wire coil.High inten-sification of mass transfer for microreactor with wire coil than plain microreactor at high temperature,leading to improves the percentage of methyl esters.
At64.7°C methanol boils rapidly and formed a large number of bubbles,which inhibits the mass transfer and causing a consider-able decrease in percentage of FAME[43].Due to existence of wire coil,as mixing intensifier in the microreactor,the interaction between the phases(liquid oil and the bubble of methanol)at this temperature increased.Thus the reduction of the methyl ester is not the same for both microreactors and lower reduction was observed for the microchannel equipped with wire coil.This caused more differences between two layouts at this temperature.
3.4.Optimization and statistical analysis
Optimization of the experimental conditions is a vital step in the transesterification process.A second-order polynomial equa-tion was developed using the Box–Behnken design to verify the factor interactions and optimize the involved parameters.The qua-dratic polynomial equation to predict the percentage of FAME using wire coil as a function of independent variables expressed by Eq.(3)as follows:FAME%¼98:5533þ1:0288X1À1:5788X2À3:7054X2
1
À4:3504X2
2
À3:3779X2
3
ð3Þ
where X1,X2and X3represent molar ratio of methanol to oil,cata-lyst concentration and reaction temperature,respectively.
The analysis of variance(ANOVA)for response surface quadratic model is given in Table3and the statistical significance of above equation was controlled by F-value.The magnitude of p-value less than0.05were chosen and indicate that the more significant model terms for prediction of the percentage of FAME[44].From Table3,it can be seen that the determination coefficient(R2)and the adjusted coefficient(R2adj.)were found96.02%and88.86%, respectively.The high value determination coefficient(0.9602) represented a good agreement between the experimental result and the predicted values and showed that the model was reliable for continuous production of biodiesel.Moreover,from the adjusted coefficient value(0.8886),it can be concluded that that the total variation of88.86%was attributed to the independent variables.
Furthermore,the quadratic equation of FAME without using wire coil is given below[38](in terms of coded factors):
FAME%¼93:5367À4:7983X2
1
À3:7183X2
2
À3:5933X2
3
ð4ÞFrom the obtained model,it can be seen that in the microreac-tor equipped with wire coil the molar ratio(methanol/oil)and cat-alyst concentration term are important in FAME%,whereas in the plain microreactor linear terms are not significant terms.Among the quadratic effect in both microreactors,all term were found to be significant on FAME%.By contrast,it can be noticed that the interaction terms were not significant on the FAME%for both microreactors.
As well as to investigate more the relationship between vari-ables and FAME%are presented using the surface plots.These plots are shown in Fig.6at constant reaction temperature of60°C, methanol/oil molar ratio(9:1)and catalyst concentration 1.2wt.%versus FAME%,respectively.
3.5.Effect of the residence time under optimal conditions
According to the literature,in many studies it was reported that the conversion rate increases with time for the transesterification reaction[14].Fig.7shows the effect of various residence times, on the percentage of methyl esters conversion.The experiments were performed for microreactors with and w
ithout wire coil at temperature of60°C,KOH concentration of1.2wt.%and the molar ratio(methanol/oil)of9.The residence times were20,26,36,60 and180s,respectively.As shown in thisfigure,the percentage of methyl esters conversion increased for microreactors with wire coil from87.9%to99%with increasing residence time reaction time from20to180s.It can be seen that the percentage of methyl esters conversion increased rapidly with increasing the residence time from20to60s then increased more slightly from60s to 180s.These results may be attributed to the incomplete transeste-rification reaction between methanol and oil in the shorter resi-dence time.A higher percentage of methyl esters in less residence time than a plain microreactor was achieved in the mic-roreactor with wire coil,which is the most significant advantage for modified microreactor over the plain one.This is obvious in Fig.7,that at residence time longer than60s,the percentage of methyl esters conversion for modified microreactor did not change significantly.
B.Aghel et al./Energy Conversion and Management84(2014)541–549545
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论