**2.3.2 The drawbacks of supercritical transesterification**

The original parameters to achieve a high transesterification conversion were a high temperature (330 – 350 ºC), high pressure (19 – 35 MPa) and high alcohol to oil molar ratio (1:40 – 1:42). Indeed, the high temperature and pressure requires both an expensive reactor and a sophisticate energy and safety management policy. As a result of the high alcohol to oil molar ratio a large energy consumption in the reactants pre-heating and recycling steps is required. Moreover, the high amount of alcohol in the biodiesel product retards the biodiesel-glycerol phase separation. Therefore, the use of those original parameters results in high capital costs, especially for the reactor and pump, being somewhat higher than the novel catalytic methods.

To increase the technical and economical feasibility of supercritical transesterification, further studies are required to reduce the energy consumption and operating parameters of this process. For example, the integration of a heating and cooling system can improve (reduce) the energy demand. The experimental techniques that have demonstrated the ability to lower the original parameters for supercritical transesterification are illustrated in sections 3.2 – 3.4.

#### **2.4 The economical feasibilities and chemical limitations of supercritical transesterification**

The economical feasibilities of supercritical transesterification, compared with the conventional homogeneous catalytic methods, have been studied by computer simulation (van Kasteren & Nisworo, 2007). These studies usually employed the original parameters for transesterification in supercritical methanol (350 ºC, 20.0 MPa and 1:42 methanol to oil molar ratio) and the general parameters for the conventional catalytic methods, such as a reaction at 60 ºC, 0.1 MPa and a methanol to oil molar ratio of 9:1. With respect to chemical limitations, supercritical transesterification is limited by the operating temperature due to the thermal cracking of the unsaturated fatty acids.

#### **2.4.1 The economical feasibilities of supercritical transesterification**

Supercritical transesterification with the original reacting parameters is economically competitive compared to the conventional catalytic method especially when low-grade feedstocks are employed (van Kasteren & Nisworo, 2007). As expected, the supercritical transesterification has a larger capital cost, due to the required reacting and pumping systems, than the conventional catalytic method, but has no additional capital and operating costs on feedstock pre-treatment, product post-treatment and waste management.

For a better economic feasibility, research into ways to reduce the high operating conditions and lower energy consumption are warranted. For example, supercritical transesterification is not economically feasible when the heating and cooling integration is not employed (Marchetti & Errazu, 2008), while it is a feasible method when heat integration and the presence of catalysts are applied (D'Ippolito et al., 2006; Glišic et al., 2009; van Kasteren & Nisworo, 2007). The addition of calcium oxide as a solid catalyst and the reduction of the alcohol to oil molar ratio significantly decreased the total energy demand and improved the economic feasibility as well (Glišic & Skala, 2009). Furthermore, the addition of propane as co-solvent also enhanced the economic feasibility of supercritical transesterification with methanol (van Kasteren & Nisworo, 2007). However, additional feasibility studies on other assisting techniques that lower the original parameters (see Sections 3.2 - 3.4) are still required.

Transesterification in Supercritical Conditions 259

2003; Kusdiana & Saka, 2004a) and the reactivity of supercritical alcohols were all reported (Warabi et al., 2004). In 2004, the first supercritical transesterification of sunflower oil with ethanol and supercritical carbon dioxide in the presence of a lipase enzyme were investigated in a batch reactor (Madras et al., 2004). However, during 2001 – 2005, the maximum alkyl ester contents were generally observed at nearly the same reaction conditions as that reported earlier by the Japanese pioneers (Kusdiana & Saka, 2001; Saka &

In 2005, carbon dioxide and propane were introduced as co-solvents to obtain milder operating parameters for the supercritical transesterification with methanol (Cao et al., 2005; Han et al., 2005). Then, the two-step supercritical process (Minami & Saka, 2006) was demonstrated to reduce those operating parameters. In the following years, various catalysts were employed to assist the supercritical transesterification to achieve the maximum alkyl esters content but at milder operating conditions (Demirbas, 2007; Wang et al., 2008; Wang et al., 2007; Wang & Yang, 2007; Yin et al., 2008b). The continuous production of biodiesel in supercritical methanol was reported in 2006 (Bunyakiat et al., 2006) (Minami & Saka, 2006) and 2007 (He et al., 2007b). Therefore, the research focus on the reduction of the elevated operating conditions and continuous process has been ongoing

In 2007, the gradual heating technique was introduced to limit or prevent thermal cracking of the unsaturated fatty acids and so prevent the reduction in the final methyl esters content obtained (He et al., 2007b). At the same time, the effect of using co-solvents to reduce the viscosity of vegetable oils was successfully investigated (Sawangkeaw et al., 2007). Supercritical transesterification in ethanol was studied in a continuous reactor in 2008 (Vieitez et al., 2008). In 2009, carbon dioxide was applied to supercritical transesterification with ethanol to reduce the operating conditions (Bertoldi et al., 2009). From 2007 to 2010, numerous additional studies, such as vapor-liquid equilibria of binary systems (Anitescu et al., 2008; Fang et al., 2008; Shimoyama et al., 2008; Shimoyama et al., 2009; Tang et al., 2006), phase behavior of the reaction mixture (Glišic & Skala, 2010; Hegel et al., 2008; Hegel et al., 2007), thermal stability of unsaturated fatty acids in supercritical methanol (Imahara et al., 2008) and process simulation and economic analysis (Busto et al., 2006; D'Ippolito et al., 2006; Deshpande et al., 2010; Diaz et al., 2009; van Kasteren & Nisworo, 2007) were reported,

leading to a better understanding of the supercritical transesterification process.

purposes and advantages that will be presented accordingly.

The co-solvents that have been used in supercritical transesterification are liquid cosolvents, such as hexane and tetrahydrofuran (THF), and gaseous co-solvents, such as propane, carbon dioxide (CO2) and nitrogen (N2). Both types of co-solvents have different

The liquid co-solvents are added into the supercritical transesterification reaction to reduce the viscosity of the vegetable oils, which might otherwise pose some pumping problems in a continuous process (Sawangkeaw et al., 2007). Since hexane is the conventional solvent for vegetable oil extraction, it could be possible to combine the supercritical transesterification after the extraction process using hexane for both. Additionally, THF improves the solubility of alcohols in the triglyceride and so forms a single phase mixture, allowing a single highpressure pump to be employed to feed the reaction mixture into the reactor. A small amount of liquid co-solvent, up to ~20% (v/v) of hexane in vegetable oil, neither affects the

Kusdiana, 2001).

since 2005.

**3.2 The addition of co-solvents** 
