A design algorithm for batch stirred tank transesterification reactors

A design algorithm for batch stirred tank transesterification reactors

Energy Conversion and Management 62 (2012) 40–46 Contents lists available at SciVerse ScienceDirect Energy Conversion and Management journal homepag...

524KB Sizes 4 Downloads 68 Views

Energy Conversion and Management 62 (2012) 40–46

Contents lists available at SciVerse ScienceDirect

Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

A design algorithm for batch stirred tank transesterification reactors C.N. Anyanwu a,⇑, C.C. Mbajiorgu b, O.U. Oparaku a, E.U. Odigboh b, U.N. Emmanuel c a

National Centre for Energy Research and Development, University of Nigeria Nsukka, Nigeria Dept. of Agricultural and Bioresources Engineering, University of Nigeria Nsukka, Nigeria c Dept. of Electronic Engineering, University of Nigeria Nsukka, Nigeria b

a r t i c l e

i n f o

Article history: Received 20 December 2011 Received in revised form 30 March 2012 Accepted 30 March 2012 Available online 25 June 2012 Keywords: Biodiesel Computer software Reactor Vegetable oil Design CI engine

a b s t r a c t A 50 L per batch, stirred tank reactor, suitable for carrying out transesterification of vegetable oils was designed and constructed. The major design assumptions included stainless steel plate thickness of 2 mm, reaction temperature of 60–65 °C and an initial/final fluid temperature of 25/70 °C. The calculated impeller Reynolds number was in the mixed regime zone of 10–104; the power number was varied between 1 and 5, while a typical propeller speed of 22.5 rev/s (or 1350 rev/min) was adopted. The limiting design conditions were maximum reactor diameter of 1.80 m, straight side height-to-diameter ratio in the range of 0.75–1.5 and minimum agitator motor power of 746 W (1 Hp). Based upon the design, a simple algorithm was developed and interpreted into Microsoft C Sharp computer programming language to enable scale up of the reactor. Performance testing of the realized reactor was carried out while using it to produce Neem oil biodiesel via base – catalyzed methanolysis, which yielded high quality fuel product. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction With global carbon emissions already in excess of the dangerously high level of 450 ppm [1], concerted efforts are now geared towards reducing the use of fossil fuels by increasing the application of renewable energy derived fuels, especially in the transport sector. Biodiesel derived from vegetable oils and fats has been demonstrated as the best suited alternative fuel for compression ignition engines. The biodiesel manufacturing process converts oils and fats into long chain mono alkyl esters, or fatty acid methyl esters (FAMEs), which are utilized as biodiesel fuel after some purification processes. Straight vegetable oils (SVOs) have been used in compression ignition (CI) engines as alternative fuels, starting with Rudolph Diesel’s successful trials in the early 19th century when he used peanut oil to power his CI engine. Research has, however shown that raw vegetable oils are capable of causing problems to the engine in the long term, owing to their high viscosity [2,3]. Transesterification of vegetable oils into biodiesel has, therefore, become a standard way of producing fuel grade products from them by reducing their viscosity to acceptable values of about 3.2 cSt. Although biodiesel has numerous advantages over fossilderived diesel fuel such as safety (high flash point, usually above 140 °C), environmental friendliness (low greenhouse gas emissions), and bio-degradability (low trace metal and sulfur contents); the major factor limiting its widespread utilization is high cost of ⇑ Corresponding author. Tel.: +234 805 1980070. E-mail address: cnasofi[email protected] (C.N. Anyanwu). 0196-8904/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.enconman.2012.03.027

production. The production cost of biodiesel is still higher than the pump price of diesel in most countries, thereby making it less attractive to the public, except with government incentives. The high production cost of biodiesel is a direct consequence of the high price of vegetable oil raw materials, most of which are edible (first generation biofuel sources) and serve as food for man. For this reason, it has been adduced that a sustainable biodiesel industry with vegetable oil starting raw material ought to be based on non-edible oils [4], often referred to as second generation biofuel sources; or even third generation biofuel resources such as algae. In addition, second generation biofuel resources such as jatropha, castor, rubber seed, and neem oils offer great potentials for the development of agriculture and light industries. Apart from high vegetable oil prices, the processing cost of biodiesel is also high, constituting 20–25% of the total production cost [5]. Arising from the need to continue research work aimed at reducing processing costs, it is necessary to study the transesterification of oils using uniform bench-scale processors to enable proper application of obtained optimal results to commercial scale plants. The main objective of the present work is to develop a computer program based on the design algorithm of a batch biodiesel reactor. Specifically, it describes the development of a C Sharp computer program for design of batch stirred tank transesterification reactors. 2. Technology and kinetics of transesterification There are five major technological means used to obtain fuel grade products from vegetable oils, namely dilution, pyrolysis,

C.N. Anyanwu et al. / Energy Conversion and Management 62 (2012) 40–46

41

Fig. 1. Schematic of acid/base transesterification process. Source: [5].

micro-emulsification, catalytic cracking and transesterification [6]. Among these, transesterification is the most commonly applied technique in view of its comparative cost advantage and convenience [7,8]. Among other methods of transesterification, such as the acid [9,10] and enzyme catalyzed processes, and (non-catalytic) supercritical methanol process [11–14]; base catalysis has found widespread industrial application owing to the fact that it produces greater yield, is simple and can be scaled up easily being a more flexible process. However, a major requirement towards the feedstock for the base catalyzed process is that it should contain less than 0.05 wt.% of water and less than 0.5 wt.% free fatty acid [15]. According to Canakci and Van Gerpen [16], transesterification will not occur when the FFA content is above 3%. A schematic flow diagram of the transesterification process is presented in Fig. 1. In the batch-scale base process, most of the steps are usually undertaken separately. The catalyst promotes an increase in solubility to allow the reaction to proceed at a reasonable rate [17], and is required because the alcohol is sparingly soluble in the oil phase. The most common catalysts used are strong mineral bases such as sodium hydroxide and potassium hydroxide. Firstly, the catalyst (usually 0.5–1.5% w/w) is mixed with the methanol (6:1 ratio to the oil, i.e. 100% excess) in a smaller container (the actual catalyst sodium or potassium methoxide is produced during this step). This mixture is then transferred into the transesterification reactor, where the vegetable oil has been charged and preheated. The reactor is covered tightly and heated to 60–65 °C for duration ranging from 30 to 120 min with agitation for optimal results. The mixture is then transfered into a settling tank (separator) with a high height-to-diameter ratio. Two phases of glycerol (lower) and biodiesel (upper) begin to emerge after a few minutes, but complete phase separation is only possible after a period of 2–8 h. The glycerol phase with a specific gravity of about 1.25 settles at the bottom, whereas the biodiesel remains on top.

3. Reactors for transesterification Both batch and flow reactors are used for transesterification of vegetable oils, but batch reactors are usually more adapted for research and testing activities, since they are more flexible and easy to control. Stirred, batch reactors for the purpose are often unbaffled tank vessels made of stainless steel plates and mounted on a suitable stand.

3.1. Chemical design of the reactor The relationship between reaction time and reactor volume (V) in the case of batch reactors is given by the equation [18]:

t ¼ NA0

Z 0

x

dx rA V

ð1Þ

where t is the time (s), NA0 the initial number of moles of reactant A, rA the reaction rate, and x is the conversion. According to some authors [19,20], the acid- and base-catalyzed transesterification of vegetable oils follow a pseudo second order chemical reaction mechanism, combined with a shunt reaction scheme at lower alcohol-to-oil ratios such as 6:1. Since the reaction time is a function of the desired conversion (x), it is possible to link a given reactor volume to degree of conversion once the reaction time is specified. A typical value of k is of the order of 0.001–0.01 d m3 mol1 s1. However, since the reactor is intended for transesterification studies at different temperatures and reaction durations, greater emphasis has been placed on the mechanical design aspects. 3.2. Mechanical design of the reactor 3.2.1. Description of the reactor The reactor for the present work was a Batch Stirred Tank (BSTR), unbaffled cylindrical vessel with conical bottom. It was heated with an electrical coil and agitated by means of a straight entering stirrer, powered by a single phase AC motor. The choice of a batch reactor is supported by literature [21], which indicates that they are more flexible and produce higher yield of product in comparison with continuous flow reactors, although the latter are often associated with lower operating costs and high throughputs. A batch reactor is often preferred for research work and in situations where the down-time is much less than the reaction time, especially for slow reactions like transesterification, which require up to 1 h. A flexible thermo-regulator capable of maintaining the temperature of a given medium within ±0.5 °C was fabricated based on the LM35 integrated circuit to enhance accurate temperature control during the transesterification process. 3.2.2. Sizing of the reactor Preliminary experiments carried out using conventional laboratory equipment (hot plate, equipped with magnetic stirrer and temperature sensor) corroborated literature data [22–24], which

42

C.N. Anyanwu et al. / Energy Conversion and Management 62 (2012) 40–46

show that the expected reaction time for the transesterification of Neem oil ranges between 0.5 and 2.0 h, depending on other reaction conditions such as temperature, alcohol-to-oil ratio and catalyst amount. In the present work, 1 h was taken as the design reaction time, whereas the reactor was assumed to work for 10 h per day.

to enable its use even in scaled-up versions of the equipment. According to Priday et al. [26], slanting propellers are not suitable for agitators requiring more than 3 Hp (2.24 kW). The design requirement is that the straight entering paddle should be mounted at a distance of DT/4 from the bottom of the reactor for effective mixing.

3.2.3. Reactor batch size A 500 L per day batch reactor operating for 10 h each day with an average reaction time (including down-time) of 1 h would have a batch volume given by the expression

3.2.6.1. Length of propeller. A distance of 10 cm was allowed between the agitator motor mounting and the top of the reactor for ease of maintenance. Therefore, the total length of the propeller shaft (Lp) was calculated as per the expression:

V bt ¼

Rcp 104

¼

500 104

¼ 0:05 ½m3 

ð2Þ

where Vbt is the batch volume, and Rcp is the capacity of the reactor, liters per day. 3.2.4. Diameter and height of reactor The reactor would not be filled to the brim due to safety considerations. It was assumed that the volume of conical section of the vessel is equal to the difference between the total volume and the effective volume. Thus,

pDT H ¼ 4V bt

ð3Þ

where DT is the diameter of vessel, m and H is the straight side height, m. According to Knudsen et al. [25], the height to diameter ratio of an unbaffled reactor should lie between 0.75 and 1.5. Therefore we assumed 1.3 times the diameter as height

Lp ¼ H þ hc  0:25DT þ 0:10 ½m

Since hc = 0.5DT the conical section was welded to the base at an angle of 135° with respect to the straight side. The outlet tap diameter was 2.5 cm (but could be up to 5.0 cm in larger tanks, such that hc was taken to be equal to the reactor radius).

Lp ¼ H þ 0:25DT þ 0:1 ½m 3.2.6.2. Diameter of propeller (Da). Tilton [27] recommended that the propeller diameter should be within (0.3–0.6)DT. In order to save materials and also operating costs since power, P = f(Da); Da was chosen as 0.4DT in the present work, such that Da = 0.148 m. 3.2.7. Power requirement of agitator motor The degree of laminarity or turbulence within the reactor was defined by the impeller Reynolds number, given by the expression [28]:

H ¼ 1:3DT Then 1:3p

DT ¼

D3T

NRe ¼

¼ 4V Bt

rffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3 4V Bt 3 4  0:05 ¼ ¼ 0:37 m 1:3p 1:3p

ð4Þ

and H = 1.3DT = 0.48 m. 3.2.5. Volume of conical section, Vc The volume of conical section (Vc) of the cylindrical vessel (where r1 and r2 are the radii of the reactor and outlet tap, respectively) was given by the expression:

Vc ¼ p

   hc  2 r 1 þ r 1 r 2 þ r 22 3

ð5Þ

Da2 Nq

l

  V c ¼ 0:17DT p 0:25D2T þ 0:0125DT þ 0:000625 ¼ 0:0078 m3

ð6Þ

With an effective volume of 50 L, the reactor should therefore be filled up to the 40 cm mark on the straight height. 3.2.6. The propeller shaft and paddle A straight-entering propeller, which is recommended for vessels with diameters less than 6 ft (1.80 m) or 1000 gallons (4 m3), was chosen for the unbaffled stainless steel tank. This was done

r1 hc r2 Fig. 2. Conical section of the reactor.

ð8Þ

where NRe is the Reynolds number (which is less than 10 for laminar and above 104 for turbulent flow regimes), Da the propeller diameter (0.15 m), N the rotational speed (22.5 rev/s), q the fluid density (900 kg/m3), and M is the fluid viscosity (0.2 Pa s). A typical value of N in single phase motors is 1350 rpm (22.5 rev/s) and was taken as design value for the agitator motor. For most vegetable oils, the density and viscosity are in the range 900 kg/m3 and 0.2 Pa s, respectively. These values were adopted in the design, leading to a Reynolds number of 2278. The power requirement of the agitator motor, P in watts was obtained from the expression [28]:

P ¼ Np qDa5 N3 In Eq. (5), hc is the height of the conical section of the reactor as shown in Fig. 2. Hence,

ð7Þ

ð9Þ

where Np is the power number, which depends on the Reynolds number and was obtained from Nomographs as given in [26]. Other symbols have the same meaning as in Eq. (8). The relationship expressed in Eq. (9) is given by plots of power number versus Reynolds as presented for different types of propellers in [27]. The Power number corresponding to a Reynolds number of 2278 for a propeller with few blades varies between 1 and 5, hence the power requirement could vary between 746.4 W and 3112 W. A 1 Hp motor was chosen for the present work as this would reduce initial and running costs. 3.2.8. Power requirement of heating coil The heat supplied by the electrical coil (Qcoil) is absorbed by the reactants, the stainless steel tank and the agitator. The balance is lost heat. Thus

Q coil ¼ Q oil þ Q steel þ Q ag þ losses

ð10Þ

where Qoil is the heat absorbed by oil; Qsteel the heat absorbed by the stainless steel vessel; Qag the heat absorbed by the agitator; and Losses is the heat losses.

43

C.N. Anyanwu et al. / Energy Conversion and Management 62 (2012) 40–46

These heat losses are often within 5–7% of the heat supplied. For the present design, a worst-case value of 10% was adopted, whereas initial and reaction temperatures of 25 °C and 70 °C, respectively for all the fluid reactants were used. Thus

0:9Q coil ¼ 45½ðC p V qÞoil þ ðC p V qÞsteel þ ðC p V qÞag 

ð11Þ

Q coil ¼ 50½ðC p V qÞoil þ ðC p V qÞsteel þ ðC p V qÞag 

ð12Þ

where Voil = Vbt is the volume of oil (reactants) = 0.05 m3, Vsteel the volume of stainless steel vessel, Vag the volume of agitator; C p oil the heat capacity of reactants (oil), C p steel the heat capacity of stainless steel = 0.5 kJ/kg K, C p ag the heat capacity of cast iron agitator = 0.6 kJ/kg K, qoil the density of oil = 900 kg/m3, qsteel the density of steel = 7800 kg/m3, and qag is the density of cast iron (agitator) = 7200 kg/m3. According to Liley et al. [28], the specific heat capacity of vegetable oils may be obtained from the expression

0:5 C p ¼ qffiffiffiffiffiffiffi þ 0:007ðt  15Þ; 15 d4

ð13Þ

where d is the density (g/cm3), Cp the Specific heat capacity (Cal/g °C), and t is the temperature (°C). 3.2.8.1. Volume of stainless steel material. The volume of material used to construct the conical section of the cylinder was taken as 5% of the entire material (as confirmed through practical calculations). Therefore, the volume of the steel material was calculated as follows:

V steel ¼ 1:05pDT Hd

ð14Þ

where DT is the reactor diameter; H the straight side height of the reactor; and d is the thickness of the steel plate (m). 3.2.8.2. Volume of cast iron agitator. The volume of material needed to fabricate the padle was taken as 10% of the volume of cast iron needed to fabricate the entire propeller. Therefore,

V CastIron ¼ 0:275pDa2 LP

ð15Þ

Hence

Q coil ¼ 50½ð900  2:15V Bt Þ þ ð7800  0:5V steel Þ þð7200  0:6V CastIron Þ Q coil ¼ 5  104 ½1:935V Bt þ 4:095pDT Hd þ 1:188pDa2 LP ; kJ

ð16Þ

The value obtained for coil energy using Eq. (16) was 5279 kJ. This quantity of heat energy must be supplied during the pre-heating period. Assuming a pre-heating period of 30 min, then the coil power amounts to 2.93 kW, including losses. The actual power rating of the heating coil used in the present design was 2.5 kW. 3.2.9. The thermoregulator A thermoregulator is a feedback mechanism used to monitor the reactor temperature and control it within a pre-set range or point. A linear type thermoregulator was fabricated based on the LM35 integrated circuit and used to control the heating element. Since the sensor circuit cannot be immersed in the liquid reactant phase, it was encased in a water-tight aluminum container and calibrated against the readings obtained from a K-type thermocouple (BK Precision), made in Taiwan. 3.2.10. Thickness and type of insulation To reduce heat losses from the reactor, it was necessary to insulate the vessel with properly selected and sized material. The insulation material was glass wool with thermal conductivity

θ1 θ2

θ3 θ4

X1

X2

X3

Fig. 3. Temperature profile of the wall of the reactor.

k = 0.03 W/mK. Assuming the average temperature of the reactor to be 70 °C, whereas the outer wall temperature should not exceed 25 °C; then, applying Fourier’s conduction heat transfer equation [25] to Fig. 3:

  dQ dt ¼ kA dh dx

ð17Þ

or

q¼

x1 k1 A1

h1  h4 þ k2xA2 2 þ k3xA3 3

where dQ/dh = q is the heat power (flux) (W), x1 = x3 the thickness of steel plate (0.002 m), x2 the thickness of insulator, m, x3 the thickness of leather covering (0.001 m), k1 the heat transfer coefficient of steel (45 w/mK), k2 the heat transfer coefficient of glass wool (0.03 W/mK), k3 the heat transfer coefficient of leather (0.03 W/ mK), A1 the Area of steel plate (m2), A2 (A3) the area of glass wool insulator (leather material) (m2), h1 the reactor temperature (70 °C), h2 the temperature at the inner metal/insulator boundary (°C), h3 the temperature at the leather/insulator boundary (°C), and h4 the outer wall temperature (25 °C). Obviously,

x2 ¼ k2 A2

  h1  h4 2x1  q k 1 A1

ð18Þ

The heat flux (power), q was obtained from the initial considerations and substituted in Eq. (17) such that x2 = 0.3 mm was found to be adequate glass wool insulation for the present design. The actual thickness of insulation used was 5.0 mm, considering availability of construction materials. The front and top views of the realized reactor are presented in Figs. 4 and 5, respectively. 4. Performance evaluation 4.1. Testing of the reactor This was carried out by using the reactor to produce biodiesel from 50 L of Neem oil. The temperature was maintained at 50 °C, using the fabricated thermoregulator. Alcohol to oil ratio of 6:1 was applied, while using 1% w/w NaOH as the catalyst. Neem oil purchased from a farm in Katsina State of Nigeria was degummed at 60 °C using 10% v/v (i.e. 5 L) of water heated to 70 °C. The mixture was then allowed to separate by gravity and the top (oil) phase was collected and subjected to acid esterification in order to reduce its FFA content. Acid esterification was carried out by heating the oil to 50 °C and adding 20% v/v (i.e. 10 L) of methanol slowly into the oil with agitation. The agitation was continued for a few minutes, after which 5% (i.e. 2.5 L) of concentrated sulfuric acid was added. Stirring was continued for another 30 min, after which the mixture was transferred into a settling tank and allowed to separate. The bottom (oil) phase was then collected into in a clean container and the free fatty acid content tested. The oil (with

44

C.N. Anyanwu et al. / Energy Conversion and Management 62 (2012) 40–46

Fig. 4. Front view of the reactor.

Fig. 5. Top view of the reactor.

a water content of 0.2%), was charged into the reactor and preheated to 50 °C (Optimum reaction temperature was determined based on reactions conducted using conventional laboratory hotplate equipped with thermoregulator.) The pre-determined quantity of NaOH was dissolved in the methanol and transferred into the reactor and the reactor was closed and maintained at 50 °C with agitation. At the end of 1 h, heating and stirring were stopped and the mixture transferred into a separating column, where it was allowed to stand for about 4 h for complete gravity separation of the biodiesel and glycerol. Thereafter, the supernatant (biodiesel) phase was decanted, washed several times with tap water until a pH value of 7.0 was obtained. The product was then subjected to quality control characterization tests, results of which are presented in Table 2. Flash point was determined using the Pensky Martens Closed cup method as per ASTM D-93 A. Kinematic viscosity was measured using the Canon Fenske Viscometer according to

ASTM ASTM D 445-04, while Acid number was determined according to ASTM D 664 method. 4.2. Testing of the C Sharp software The software was tested while using it to carry out design calculations for different sizes of batch stirred tank reactors with similar features to the one already constructed. Reactors with batch size of between 2 L and 10,000 L were sized using the programme. These results are presented in Appendices A–E. 5. Results and discussion The fabricated 50 L batch, stirred tank, reactor was used to process neem oil biodiesel, after acid esterification of the feedstock to

C.N. Anyanwu et al. / Energy Conversion and Management 62 (2012) 40–46 Table 1 Properties of the pretreated neem oil. S/No

Property

Value

Units

1 2 3

Water content Density Free fatty acid

0.20 881 0.13

% kg/m3 %

Nsukka for funding support provided for this research work. Contributions made by Prof. C.A. Nwadinigwe, technical staff of N.C.E.R.D. and colleagues in the Department of Agricultural and Bioresources Engineering are highly acknowledged.

Appendix A A.1. Computation details Table 2 Determined properties of the biodiesel.

*

S/No

Property

Values

ASTM limits*

Units

1 2 3 4 5 6

Flash point Kinematic viscosity Water and sediments Acid number Free glycerol Total glycerol

168 3.7 0.03 0.24 0.012 0.016

>130 1.9–6.0 <0.05 0.50 max 0.020 0.024

°C mm2/s % mg KOH/g wt.% wt.%

As per ASTM D6751 requirements. Source: [17].

reduce its FFA content to 0.13%. The methyl ester yield obtained at the end of washing was 84.6%, which is comparable with results reported elsewhere [24]. The properties of the pretreated neem oil are presented in Table 1. Values obtained for all the parameters of the produced methyl ester (Table 2) were within ASTM D6751 standard stipulated limits, an indication that the product can be used as fuel grade biodiesel. It was observed that agitation alone required 0.75 kW h during the reaction, compared to heating that consumed about 1.96 kW h since the total heating power was for pre-heating (which lasted for 22.5 min) and reaction (24.6 min). Although the energy required for agitation was only 27.6% of the processing costs (excluding pre-treatment), it could be reduced further if a 0.5 Hp (374 W) centrifugal pump had been employed to circulate the reactants through the reactor. Using the developed C Sharp software for reactor design is a time-saving measure, which also reduces the incidence of human errors in the design calculations. Tests carried out using the software also indicated that it can be used by professionals who have little or no in-depth knowledge of reactor design practice. Nonetheless, it is clear that reactor scale-up is not a purely mechanical design problem, considering the fact that the mechanical aspects have a great influence on diffusion, heat transfer and chemical kinetics; which determine the reactor’s overall performance. Adeyemi et al. [29], who studied the effect of certain variables on FAME yield concluded that there is high correlation between FAME yield, peak time and reaction temperature; although wide impeller speed variations for Rushton and Elephant ear types gave rise to only little difference in FAME yield between 89% and 94%. Future research work would investigate the influence of reactor size and degree of agitation on transesterification kinetics and quality of final biodiesel product. 6. Conclusion A design algorithm for batch stirred tank transesterification reactors was developed. On the basis of the algorithm, a generic MS C Sharp programme was developed and tested by using it to scale up the designed reactor. Performance evaluation of the constructed reactor, which was also carried out, showed good results. Acknowledgments The authors are grateful to the Management of the National Centre for Energy Research and Development, University of Nigeria

45

Reactor capacity, Rcp: 20 L. Batch size of reactor, Vbt: 0.002 m3. Diameter of vessel, DT: 0.1221 m. Straight side height, H: 0.1587 m. Volume of conical section, Vc: 0.0004 m3. Safe loading mark, h: 0.0342 m. Length of propeller: 0.3892 m. Reynolds number, NRe: 241.1208. Power requirement of the agitator motor: 2.8368 W. Thickness of stainless steel sheet: 0.002 m. Diameter of propeller, Da: 0.0366 m. Length of propeller, Lp: 0.3892 m. Pre-heating time, t: 30 min. Qcoil: 195.9459 J. Power requirement of heating coil: 0.1089 kW. Appendix B B.1. Computation details Reactor capacity, Rcp: 500 L. Batch size of reactor, Vbt: 0.05 m3. Diameter of vessel, DT: 0.3569 m. Straight side height, H: 0.464 m. Volume of conical section, Vc: 0.007 m3. Safe loading mark, h: 0.07 m. Length of propeller: 0.7532 m. Reynolds number, NRe: 2064.6738. Power requirement of the agitator motor: 608.6462 W. Thickness of stainless steel sheet: 0.002 m. Diameter of propeller, Da: 0.1071 m. Length of propeller, Lp: 0.7532 m. Pre-heating time, t: 30 min. Qcoil: 4874.0195 J. Power requirement of heating coil: 2.7078 kW. Appendix C C.1. Computation details Reactor capacity, Rcp: 5000 L. Batch size of reactor, Vbt: 0.5 m3. Diameter of vessel DT: 0.769 m. Straight side height, H: 0.9997 m. Volume of conical section, Vc: 0.0649 m3. Safe loading mark, h: 0.1397 m. Length of propeller: 1.392 m. Reynolds number, NRe: 9580.0482. Power requirement of the agitator motor: 28226.4019 W. Thickness of stainless steel sheet: 0.002 m. Diameter of propeller, Da: 0.2307 m. Length of propeller, Lp: 1.392 m. Pre-heating time, t: 30 min. Qcoil: 48671.3591 J. Power requirement of heating coil: 27.0396 kW.

46

C.N. Anyanwu et al. / Energy Conversion and Management 62 (2012) 40–46

Appendix D D.1. Computation details Reactor capacity, Rcp: 10,000 L. Batch size of reactor, Vbt: 1 m3. Diameter of vessel, DT: 0.9689 m. Straight side height, H: 1.2596 m. Volume of conical section, Vc: 0.128 m3. Safe loading mark, h: 0.1736. Length of propeller: 1.7018 m. Reynolds number, NRe: 15211.1682. Power requirement of the agitator motor: 89669.0786 W. Thickness of stainless steel sheet: 0.020 m. Diameter of propeller, Da: 0.2907 m. Length of propeller, Lp: 1.7018 m. Pre-heating time, t: 30 min. Qcoil: 97318.2696 J. Power requirement of heating coil: 54.0657 kW. Appendix E E.1. Computation details Reactor capacity, Rcp: 20,000 L. Batch size of reactor, Vbt: 2 m3. Diameter of vessel, DT: 1.2207 m. Straight side height, H: 1.5869 m. Volume of conical section, Vc: 0.2532 m3. Safe loading mark, h: 0.2163 m. Length of propeller: 2.0921 m. Reynolds number, NRe: 24141.7351. Power requirement of the agitator motor: 284549.2772 W. Thickness of stainless steel sheet: 0.020 m. Diameter of propeller, Da: 0.3662 m. Length of propeller, Lp: 2.0921 m. Pre-heating time, t: 30 min. Qcoil: 194597.1521 J. Power requirement of heating coil: 108.1095 kW.

References [1] Schenk PM, Thomas-Hall SR, Stephens E, Marx UC, Mussgnug JH, Posten C. Second generation biofuels: high-efficiency microalgae for biodiesel production. Bioenergy Resour 2008;1:20–43. [2] Schuchardt U, Sercheli R, Vargas RM. Transesterification of vegetable oils: a review. J Braz Chem Soc 1998;9(3):1–27. [3] Nwafor OMl. Emission characteristics of diesel engine running on vegetable oil with elevated fuel inlet temperature. Int J Biomass Bioenergy 2004;27:507–11.

[4] Balat M. Potential alternatives to edible oils for biodiesel production – a review of current work. Energy Convers Manage 2011;52(2):1479–92. [5] Tyson KS. Biodiesel handling and use guidelines. US Department of Energy; 2004. [6] Yusuf NNAN, Kamarudin SK, Yaakub Z. Overview of the current trends in biodiesel production. Energy Convers Manage 2011;52(7):2741–51. [7] Anyanwu CN, Sharma VK, Braccio G, Nanna F, Akubuo CO. Palm oil (Elaeis guiniensis) plantations: a potential biomass resource for biodiesel production in Nigeria. Int Energy J 2005;6(2):69–82. [8] Demirbas A. Biodiesel production from vegetable oils by supercritical methanol. J Sci Indust Res 2005:858–65. [9] Miao X, Li R, Yao H. Effective acid-catalyzed transesterification for biodiesel production. Energy Convers Manage 2009;50(10):2680–4. [10] Samart C, Chaiya C, Reubroychareon P. Biodiesel production by methanolysis of soybean oil using calcium supported on mesophorus silica catalyst. Energy Convers Manage 2010;51(7):1428–31. [11] Baroutian S, Aroua MK, Abdul Aziz AR, Sulaiman NMN. TiO2/Al2O3 membrane reactor equipment with a methanol recovery unit to produce palm oil biodiesel. Int J Energy Res 2012;36::120–129. John Wiley. [12] Demirbas A. Recent trends in biodiesel fuels. Energy Convers Manage 2009;50(1):14–34. [13] Hanh HD, Dong NT, Starvarche C, Okitsu K, Maeda Y, Nishimura R. Methanolysis of triolein by low frequency ultrasonic irradiation. Energy Convers Manage 2008;49(2):276–80. [14] Demirbas A. Biodiesel production via non-catalytic SCF methanol and biodiesel fuel characteristics. Energy Convers Manage 2006;47(15–16):2271–82. [15] Marchetti JM. The effect of economic variables over biodiesel production plant. Energy Convers Manage 2011;52(10):3227–33. [16] Canakci M, Van Gerpen J. Biodiesel production from oils and fats with high free fatty acids. Trans ASAE 2001;44:1429–36. [17] Van Gerpen J, Shanks B, Pruszko R, Clements D, Knothe G. Biodiesel analytical methods August 2002–January 2004. National Renewable Energy Laboratory Subcontractor Report NREL/SR-510-36240, Colorado, USA; July, 2004. [18] Fogler HS. Elements of chemical reaction engineering. 4th ed. Prentice Hall Int’l Series in Chemical Engineering; 2006. [19] Noureddini H, Zhu D. Kinetics of transesterification of soybean oil. J Am Oil Chem Assoc 1997;74:1457–63. [20] Freedman B, Kwoiek WF, Pryde EH. Quantitation in the analysis of transesterified soybean oil by capillary gas chromatography. J Am Oil Chem Soc 1989;63:1370–5. [21] Lin K, Van Ness HC, Abbott MM. Reaction kinetics, reactor design and thermodynamics. In: Perry RH, Chilton CH, editors. Chemical engineers’ handbook. McGraw-Hill; 1973 [Chapter 4]. [22] Sekhar MC, Mamilla VR, Mallikarjuan MV, Reddy KVK. Biodiesel production from neem oil. Int J Eng Stud 2009;1:295–302. [23] Galadima A, Garba ZN. Catalytic synthesis of ethyl ester from some common oils. Sci World J 2009;4:1–5. [24] Eevera T, Rajendran K, Saradha S. Biodiesel production process optimization and characterization to assess the suitability of the product for various environmental conditions. Renew Energy 2009;34:762–5. [25] Knudsen JG, Hottel HC, Sarofim AF, Wankart PC, Kneabel KS. Heat and mass transfer. In: Perry RH, Green DW, Maloney JO, editors. Perry’s chemical engineers’ handbook. NY: McGraw-Hill; 1999 [Section 5]. [26] Priday G, Silverblatt CE, Slottee JS, Smith JC, Todd DB. Liquid–solid operations and equipment. In: Perry RH, Green DW, Maloney JO, editors. Perry’s chemical engineers’ handbook. NY: McGraw-Hill Book Company; 1999 [Chapter 18]. [27] Tilton JN. Fluid and particle dynamics. In: Perry RH, Green DW, Maloney JO, editors. Perry’s chemical engineers’ handbook. NY: McGraw-Hill; 1999 [Section 6]. [28] Liley PE, Thomson GH, Friend DG, Daubert TE, Buck E. Physical and chemical data. In: Perry RH, Green DW, Maloney JO, editors. Perry’s chemical engineers’ handbook. NY: McGraw-Hill; 1999 [Section 2]. [29] Adeyemi NA, Mohiuddin A, Jameel T. Waste cooking oil transesterification: influence of impeller type, temperature, speed and bottom clearance on FAME yield. Afr J Biotechnol 2011;10(44):8914–29.