Simulation analysis of methanol flash distillation circulation process in biodiesel production with supercritical method

Abstract

High methanol-to-oil ratio is required to obtain a high conversion of oil for the production of biodiesel with supercritical methanol. Recovering the methanol of a stream issuing from a transesterification supercritical reactor by flash distillation instead of evaporation was analyzed. The one-stage and two-stage flash distillation processes were presented and compared. The difference of the recovery percentage of methanol of the above two flash processes is less than 0.5% and the methanol concentration in the vapor for the one-stage process decreases rapidly when feed temperature increases. The process in which the product of transesterification of soybean oil with supercritical methanol is cooled to an appropriate temperature (about 240°C) first and then flashed was put forward. The effect of cooling temperature, feed pressure and flash pressure on methanol concentration and recovery percentage was investigated. According to this study, when the feed pressure range is 15–30 MPa, the flash pressure equals 0.4 MPa, and cooling temperature range is 240°C–250°C, the recovery percentage of methanol is not less than 85%, and the concentration of the vapor in mass fraction of methanol is approximately 99%. Thus, the vapor leaving the flash tank can be directly circulated to the transesterification reactor.

Keywords

supercritical methanol / biodiesel / recovery / flash distillation

Introduction

Vegetable oil (m)ethyl esters, commonly referred to as “biodiesel” are prominent candidates as alternative diesel fuels. Biodiesel has become more attractive recently because of its environmental benefits. Biodiesel is also better than diesel fuel in terms of sulfur content, ash point, aromatic content and biodegradability. Researchers have developed various methods of preparing biodiesel such as the direct blending of oils, micro-emulsification, pyrolysis, transesterification with the help of basic, acidic and enzymatic catalysts, supercritical methanol transesterification, simultaneous reaction of transesterification and cracking, and superheated methanol vapor treatment [1-3]. Supercritical methanol transesterification method is a new technology for producing biodiesel. It does not use a catalyst, but has a higher reaction rate, a shorter reaction time (the yield of conversion increases to 95% in 10 min [4]), simpler after-treatment of the product, and so on. Supercritical methanol method has become one of the research topics in the biodiesel field. However, this method requires a high mole ratio of alcohol to oil (from 20∶1 to 40∶1) in order to obtain a high conversion of oil. The unreacted methanol should be removed from the reactor product. In the most common method (called cooling-evaporating method), the reactor product is cooled to normal temperature first, and is then laminated and evaporated. Moreover, since heat recovery is not considered, this method consumes much energy. Therefore, the research on methanol recovery and heat utilization becomes essential. In order to reduce the methanol-to-oil ratio, the reaction temperature and the pressure of the supercritical method, D’Ippolito et al. [5] proposed a two-step, supercritical biodiesel production process with heat recovery composed of two supercritical reactors operating in series. An intermediate step of glycerol removal between the two reaction steps is introduced. Double tube heat exchangers before the supercritical reactors allow the preheating of the reacting mixture by the stream exiting the reactor. Adiabatic flash drums downstream of the reactors evaporate the unreacted methanol. This process enables the use of a low methanol-to-oil ratio (10-15), the decrease of the reaction temperature to 275°C-290°C and the reduction of the working pressure to about 10-11 MPa. Liu et al. [6,7] fulfilled the lab and bench-scale experiments of the one-step, supercritical biodiesel production process. Considering high temperature and pressure of the reactor product, the scheme of methanol recovery by flash distillation instead of evaporation will be discussed in this paper. It can be expected to obtain a high purity of methanol with lower energy cost. This study will focus on recovering methanol from the stream exiting from the transesterification reactor with soybean oil and supercritical methanol as reactants in the flash distillation process.

Prediction of critical properties of soybean oil and soybean methyl ester

Soybean oil is a mixture of various triglycerides. The acids in the triglycerides can be saturated and /or unsaturated fatty acids. Correspondingly, soybean methyl ester (SME) is a mixture of unsaturated and saturated fatty acid methyl esters. The compositions of soybean oil and SME are listed in Tables 1-2 [8].

For simplicity, soybean oil and SME are considered as a single component in simulation. The virtual structure of soybean oil can be obtained as follows. First, soybean oil is considered as a mixture of several triglycerides. Second, calculate the molar fractions of the triglycerides. Finally, determine the number of groups CH₂, CH=CH and CH₃ in soybean oil by the linear mixing rule of molar fraction. Similarly, the virtual structure of SME can be obtained.

In simulation, the critical properties and acentic factors of soybean oil and SME were obtained through estimation. In various methods of estimating critical properties, group mixing method is reliable and comparably generally used. Constantinous-Gani (C-G) method presented in 1994 [9] was used to estimate the critical temperature, pressure and normal point. The estimated results are listed in Table 3. According to Ref. [10], the normal boiling point ranges are 616.3-625.3 K. The estimated results of this study can be reliable. Of the three methods of estimating acentic factor [9], namely, Edmister, Lee-Kesler and Chen equations, the third has the least error. The estimated acentic factors of soybean oil and SME with Chen equation are also listed in Table 3.

Methanol recovery process with flash distillation

In this study, the stream issuing from the transesterification reactor is flashed directly. Two possible flash-distillation processes are put forward (Figs. 1-2). The numbers 1, 2, 3, 4, 5, 6, 7 in the square frames represent the number of stream.

Mathematical model of adiabatic flash distillation

Component i balance:

(1)q_{\mathrm{f}}{z}_i=q_{\mathrm{v}}{y}_i+q_{\mathrm{l}}{x}_i.

Phase equilibrium equation:

(2)y_i=K_i{x}_i\mathrm{.}

Enthalpy balance:

(3)q_{\mathrm{f}}{H}_{\mathrm{f}}=q_{\mathrm{v}}{H}_{\mathrm{v}}+q_{\mathrm{l}}{H}_{\mathrm{l}}.

Dew point equation:

(4){\displaystyle \sum y_i}=\mathrm{1.}

Bubble point equation:

(5){\displaystyle \sum x_i}=\mathrm{1.}

In Eqs. (1)-(5), q_f, q_v and q_l are the flowrates of the feed, the vapor stream and the liquid stream, respectively; z_i, y_i, and x_i are the concentrations of component i of the feed, the vapor stream and the liquid stream, respectively; K_i is the phase equilibrium constant of component i; H_f, H_v, and H_l represent the enthalpies of the feed, the vapor stream, and the liquid stream, respectively.

The phase equilibrium constant K is calculated by PSRK equation [11] and enthalpy H by mixing model. The enthalpy of methanol is calculated by latent-heat model and that of soybean oil, SME, glycerol by PSRK model.

Results and discussion

Determination of flash distillation process

According to studies on the production of biodiesel with supercritical methanol, the molar ratio of methanol to vegetable oil is 20∶1-40∶1, the reaction temperature is 250°C-400°C, the reaction pressure is 10-35 MPa, and the conversion of soybean oil is approximately 95%. Considering the pressure drop of the flow of fluid, the low pressure value of the final flash tank should be 0.3-0.5 MPa. For comparison of the one-stage flash process with the two-stage, the final flash pressure is set at 0.4 MPa and the pressure of flash tank 2 (see Fig.2) at 2 MPa. The feed composition is determined according to the molar ratio of methanol to soybean oil 30∶1 and the conversion of oil 95%, namely, the content of soybean oil is 2.34%, that of methanol is 47.30%, that of SME is 45.61%, and that of glycerol is 4.75% in the mass fraction. The calculated results are shown in Figs. 3-5. The symbol η denotes the total recovery percentage of methanol, which equals the quantity of methanol in the vapor stream divided by that in the feed, the symbol w₃ represents the mass fraction of the methanol in stream 3 ( see Figs. 1 and 2), and the symbol w₆ represents the mass fraction of the methanol in stream 6 ( see Fig.2).

Figures 3-5 show that under the same feed conditions, the difference of the recovery percent of methanol between the one-stage and two-stage flash distillation processes is less than 0.5%, and the concentration difference of methanol of stream 3 issuing from flash tank 2 for two processes is less than 2% in mass fraction. The concentration difference of methanol between stream 3 issuing from flash tank 2 (see Fig. 1) and stream 6 from flash tank 4 (see Fig. 2) increases rapidly when feed temperature increases. However, the capital and operation cost of the two-stage process is more than that of the one-stage. Therefore, the one-stage flash process is chosen. The results also show that the methanol concentration of the vapor decreases rapidly when feed temperature increases. For example, w₃ reduces from 97.89% to 73.47% when feed temperature rises from 260°C to 360°C at the feed pressure of 15 MPa. The reason for this is that more and more glycerol and SME enter the vapor phase. Based on this, an idea that the feed is cooled to an appropriate temperature instead of normal temperature and followed by flash distillation is put forward in order to adapt the one-stage flash process to various feed conditions and obtain a high purity of methanol (Fig. 6).

Effect of cooling temperature and feed pressure on methanol recovery

In order to investigate the effect of cooling temperature and feed pressure on methanol recovery, the flash tank pressure is set to 0.4 MPa, the feed (stream 1 in Fig. 6) composition is the same as that in section 5.1. The calculated results of the feed pressure of 15 MPa, 20 MPa, 25 MPa and 30 MPa at a cooling temperature of 210°C-260°C are shown in Fig. 7, where w₄ represents the mass fraction of the methanol in stream 4. The results show that for a given pressure and composition of the feed, the concentration of the vapor in the mass fraction of methanol increases slowly and η decreases rapidly when cooling temperature decreases. For a given feed composition and cooling temperature, the quality and quantity of methanol recovery are slightly affected by the feed pressure. As for the investigated conditions, the optimum cooling temperature is 240°C-250°C. Under this condition, the concentration of the vapor in the mass fraction is approximately 99% and the recovery percentage is not less than 85%. In industry practice, the optimum cooling temperature should be selected in the light of the feed temperature and the removed heat needed for the cooler in order to utilize the heat effectively and improve the quality and quantity of methanol recovery.

Effect of flash pressure on methanol recovery

In order to investigate the effect of flash pressure on methanol recovery, the cooling temperature is set to 240°C, the feed composition (stream 1 in Fig. 6) is the same as that in section 5.1. The calculated results under the feed pressure of 15 MPa and 25 MPa are shown in Fig. 8. The results show that as for certain pressure and composition of feed, both the concentration of the vapor in the mass fraction of methanol and recovery percentage of methanol increase when flash pressure decreases. In order to obtain the quality and quantity of recovering methanol, the flash pressure should be as low as possible.

Conclusions

Biodiesel production with supercritical methanol method requires high mole ratio of methanol to oil. The conventional one-step supercritical method with no heat recovery consumes much energy. This study focuses in detail on simulation analysis of methanol recovering from the transesterification reaction product (which was taken as the feed of flash distillation) by flash distillation. The following conclusions were obtained through the simulation.

1) The recovery percentage of methanol for the one-stage and flash distillation processes is close to that for the two-stage under the same feed conditions. The capital and operation cost of the one-stage process is less than that of the two-stage. Therefore, the one-stage process is chosen.

2) The concentration of methanol recovered from the one-stage process decreases rapidly when feed temperature increases. Based on this, the appropriate cooling-flashing process is proposed in order to obtain a high concentration of methanol for various feed conditions.

3) As for the appropriate cooling-flashing process, the concentration of methanol recovered is slightly affected by cooling temperature. The recovery percentage of methanol decreases rapidly when cooling temperature decreases. Both the concentration of methanol recovered and recovery percentage of methanol increase when flash pressure decreases. According to this study, when the pressure of the feed is 15-30 MPa, the flash pressure is 0.4 MPa, and cooling temperature is 240°C-250°C, the recovery percentage of methanol is not less than 85%, and the concentration of the vapor in the mass fraction of methanol is approximately 99%. Therefore, the vapor leaving the flash tank can be circulated to the transesterification reactor directly.

References

[1] Tang Shenglan, Zhang Jianan, Wu Haitang, Li Chun. Research advances in preparation of biodiesel fuel using supercritical techniques. Modern Chemical Industry, 2006, 26(Suppl.2): 391-394 (in Chinese)

[2] Demirbas A. Recent developments in biodiesel fuels. International Journal of Green Energy, 2007, 4(1): 15-26

[3] Nabetani H. Development of non-catalytic reaction processes for production of biodiesel fuel. Journal of the Japan Institute of Energy, 2007, 86(6): 371-374

[4] Demirbas A. Biodiesel production via non-catalytic SCF method and biodiesel fuel characteristics. Energy Conversion and Management, 2006, 47(15,16): 2271-2282

[5] D'Ippolito S A, Yori J C, Iturria M E, Pieck C L, Vera C R. Analysis of a two-step, noncatalystic, supercritical biodiesel production process with heat recovery. Energy and Fuels, 2007, 21(1): 339-346

[6] Liu Qidong, Wang Cunwen, Wang Weiguo, Wu Yuanxin, Zhang Junfeng, Chi Ruan, Li Shirong, Ding Yigang. The primary study of continuous preparation of biodiesel under supercritical condition. Chemistry & Bioengineering, 2006, 23(4): 26-28 (in Chinese)

[7] Wang Tielin, Wang Cunwen, Zhao Anmin, Wang Weiguo, Zhang Junfeng. Bench-scale research for continuous preparation of bio-diesel with supercritical methanol in a single tubular reactor. Modern Chemical Industry, 2008, 28(12): 71-75 (in Chinese)

[8] Wu Miaoxin, Wu Guoying, Han Ying, Zhang Peidi. Fatty acid composition of six edible vegetable oils and their biodiesel fuels. Chinese Fats, 2003, 28(12): 65-67 (in Chinese)

[9] Dong Xinfa. Principles of Physical Property Estimation and Calculation with Computer. Beijing: Chemical Industry Press, 2007, 44-45, 151-159

[10] Yuan W, Hansen A C, Zhang Q. Vapor pressure and normal boiling point predictions for pure methyl esters and biodiesel fuels. Fuel, 2005, 84(7,8): 943-950

[11] Horstmann S, Fischer K, Gmehling J. PSRK group contribution equation of state: Revision and extension III. Fluid Phase Equilibria, 2000, 167(2): 173-186

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg