Elsevier

Fuel

Volume 255, 1 November 2019, 115781
Fuel

Full Length Article
Numerical modelling of free energy for methanol and water mixtures for biodiesel production

https://doi.org/10.1016/j.fuel.2019.115781Get rights and content

Highlights

Abstract

As the demand for energy is ever increasing, biodiesel has become a major alternative to fossil fuels, primarily due to the latter’s depletion rate. In methanol-water rectification, a vital part of the biodiesel production process, distillation is the preferred technique to recover methanol from water. This study aims to investigate the excess Gibbs free energy (ΔG) behavior on methanol in a water mixture, for the binary system. The study was conducted using the GROningen MAchine for Chemical Simulations (GROMACS©). The effect of several important parameters, namely, sub-atmospheric pressure, temperature and methanol concentration, were investigated on the excess Gibbs free energy of methanol in water. The simulation results were partially validated, since the experimental data was limited. The results demonstrate that the excess Gibbs free energy of methanol in a water mixture positively increases with an increase in methanol concentration, until it reaches a threshold concentration of 0.5. The positive excess Gibbs free energy of the mixture system indicates that the process is not spontaneous and requires additional energy to occur. Increasing the temperature directly increases the excess Gibbs free energy of the system. Increasing the pressure does not have a significant increase on the excess Gibbs free energy. The simulation results are comparable with the experimental results; as validated by the Pearson correlation coefficient (ρ) = +0.9982. This study is beneficial to predict the thermodynamic behavior of methanol in a water system, in order to contribute to biodiesel production process optimization.

Introduction

According to the U.S. Energy Information Administration (EIA), by the year 2040, the world energy consumption will increase up to 30% [1]. The EIA also forecasts that by the year 2040, fossil fuel production will decline, and will be replaced by renewable sources of energy [1]. Global awareness of environmental pollutions and energy issues have stimulated many researchers toward finding alternatives for fossil fuels. One of the most promising alternative fossil fuel sources is biodiesel for several reasons. Some of the biodiesel’s promising properties are high biodegradability, minimum toxicity and the ability to replace diesel fuel across many different applications without any major modifications. It has almost zero emission of sulfates, aromatic compounds and other environmentally destructive chemical substances, and has a small net contribution of carbon dioxide (CO2) [2], [3], [4]. Recent studies have shown that biodiesel is able to give a similar performance compared to diesel fuels [5], [6]. Many countries have explored the potential of biodiesel as an alternative. The five leading countries pursuing this interest is Malaysia, Indonesia, Argentina, United State of America, and Brazil [2]. Based on EIA statistics, the global biodiesel production has gradually increased from 15,000 bbls/day in the year 2000, to 289,000 bbls/day in the year 2008, and is expected to increase in the near future [1].

Many research works have been conducted to produce vegetable oil derivatives which have similar chemical and physical properties, as well as comparable performance to that of diesel fuel. However, these are not without their own challenges. Common problems which can be faced when dealing with vegetable oil is its low volatility, high viscosity and its polyunsaturated characteristics [2]. The most commonly used method to overcome these problems is through the transesterification process [2].

The process of biodiesel production starts with the mixing of alcohol (typically methanol), oil, and a catalyst in the reactor [7]. During the biodiesel production process, the excess methanol in the system tends to absorb water. This water will be removed during the methanol-water rectification step in the distillation column. The purified methanol will then be recycled back for other batches of biodiesel production.

To proceed with the next step, which is excess methanol removal, the separation of glycerol from methyl esters need to be fully achieved to accommodate the reversion of the transesterification reaction. After methanol removal, methyl ester will go through a neutralization step, where acid will be added. In this step, acid will be used to neutralize any residual catalyst and split any soap that is formed into free fatty acids and salts [7]. The methyl ester will then enter a water washing step, for the removal of any remaining soap, free glycerol, catalysts and salt which formed due to the reaction between soap and acid. Any remaining water that has been used for the water wash step will be removed in the vacuum flash process. The total percentage of the glycerol stream that leaves the separation process is only about 50%, while the rest are mostly catalyst, soap and some excess methanol. It will then enter the methanol removal step, where methanol is removed by the vacuum flash or the evaporator.

It is important to understand the recovery of methanol in water mixtures during the distillation process. Excess Gibbs free energy (ΔG) is one of the significant thermodynamic equations used to quantify and understand the phase formation, as well as the biological and the chemical behavior of the mixture [8]. In general, the activity coefficients can be used to understand the equilibrium condition of the system. The excess Gibbs free energy is related to the activity coefficients [9]. The relation to the activity coefficients and the excess Gibbs free energy is shown in Eq. (1).ΔGRT=lnγwhere γ is the activity coefficient, ΔG is the excess Gibbs free energy (kJ/mol), R is the universal gas constant and T is the temperature.

The excess Gibbs free energy is a spontaneity measurement of the system’s thermodynamic departure from the ideal behavior. The ideal behavior can be understood when the mixture does not have a chemical reaction, is repulsive and attracts forces amongst other molecules [10]. Hence, decreasing the excess Gibbs energy or the more negative values indicates that the process is thermodynamically favored [8].

Free licensed software GROningen MAchine for Chemical Simulations Molecular Dynamic (GROMACS© MD) was used to investigate the excess Gibbs free energy of methanol in the water mixture for the binary system. Pressure, temperature and methanol concentration were varied to study their effects on the excess Gibbs free energy. The molecular dynamic simulation has been proven by many researchers to give comparable overall agreement with the experimental work. In recent years, there has been a considerable shift from experimental work to molecular dynamic simulation. However, extensive research is still required until molecular dynamic simulation becomes a norm. This can better answer arguments, because experimental results are more reliable as it considers real-life environments [8]. The simulation results in this study are validated with the experimental results from Soujanya et al. (2010). These simulation results are partially validated since the experimental data is limited.

The study started with the simulation set up, by selecting suitable parameters and inputs for the design of experiments. It was then followed by building the pre-equilibrated simulation box of methanol and water using the GROMACS© simulation engine. The molecules were packed in the simulation box using PACKMOL, where the dimension was set at 30 Å on each side. Following the simulation box packing, the simulation was run to allow the molecular interaction in GROMACS© and to confirm the ΔG. If the simulation was functional and ran without any error, the parametric study was conducted by varying the methanol concentrations to study the effect of methanol concentration on the ΔG. The simulation results were partially validated since the experimental data was limited. The GROMACS© simulation was then conducted to calculate the excess Gibbs free energy of the mixture system, while the component’s molecular structure was built using PyMOL. Analysis of Variance, (ANOVA) was used in this work to investigate the effect of the parameters of interest on the excess Gibbs free energy of methanol and water system. In this simulation study, the null hypothesis (Ho) postulates that these parameters do not impact the excess Gibbs free energy. On the other hand, the alternative hypothesis (H1) represents parameters which statistically impact the excess Gibbs free energy.

The current study suggests that molecular dynamic simulation can be employed with high confidence and at lower cost to replace laboratory experimental measurements of the excess Gibbs free energy of methanol and water mixtures. In addition to that, the simulation work can be conducted at extreme pressures and temperature conditions, which are the main limitations of laboratory experimental study. One of the novelties of this study is the positive discovery of excess Gibbs free energies for methanol in water for the binary system, which indicates the not spontaneous mixing reaction in the system. Results attained from ANOVA indicate that the excess Gibbs free energy is significantly affected by methanol concentration and temperature, whereas the methanol pressure in the water system does not pose a significant effect. The results propose that at a specific temperature and pressure, and a higher methanol concentration results in higher excess Gibbs free energy of methanol in the water mixtures until it reaches XMeOH = 0.5. Further increase in the methanol concentration will decrease the excess Gibbs free energy. The study also suggests that one of the cost-effective methods to increase methanol’s recovery during the distillation process in biodiesel production is by reducing the excess Gibbs free energy. The methanol’s recovery improvement can be performed by finding the optimum pressure, temperature and methanol concentration of methanol in the water system. For the range of parameter values that were studied, the optimum pressure was 29.38 kPa, the optimum temperature was 313.3 K, and the optimum methanol concentration stood at 0.3.

Section snippets

Simulation setup and input parameters

The simulation setup and the range of parameters values to study the excess Gibbs free energy of methanol and water in a binary system are tabulated in Table 1. The range of temperature and pressure used in the simulation were taken from the experimental data conducted by Soujanya et al. (2010). The experimental data have been obtained at sub-atmospheric pressure condition. The methanol concentration range represents the transition between the water-rich phase to the methanol rich environment.

Effect of temperature and pressure on the excess Gibbs free energy of methanol in water

Based on the simulated data, the excess Gibbs free energy for methanol in water for the binary system were positive, which indicated the unspontaneity for the recovery of methanol from water. For the better understanding of the excess Gibbs free energy of the binary system, it is preferred to have an insight on the thermodynamic behavior by separately looking into the individual molecules relative to the pure liquids.

Pascal and Goddard (2012) showed that the excess Gibbs free energy of the

Conclusion

The main objectives of this work were to study the effects of temperature, pressure and methanol concentration on the excess Gibbs free energy of methanol in water. This study proposed the following:

  • 1.

    Excess Gibbs free energies for methanol in water of binary system were positive, indicating the unspontaneity of the reaction in the system.

  • 2.

    It is observed that the excess Gibbs free energy of the system increases with the increase in temperature.

  • 3.

    The increase in pressure affects the excess Gibbs free

Acknowledgment

The authors wish to acknowledge the Petroleum Engineering Department of Universiti Teknologi PETRONAS for their support during the course of this research work.

Cited by (0)

View full text