NEDO Biodiesel Production Process by Supercritical Methanol Technologies

NEDO Biodiesel Production Process by Supercritical Methanol Technologies Shiro Saka * Graduate School of Energy Science, Kyoto University, Kyoto, Japan Abstract: Biodiesel fuel is expected to contribute to mitigating our environmental and energy problems particularly in transportation sector. At present, a commercial process for biodiesel production involves the use of alkaline catalyst. Therefore, sophisticated purification steps are necessary to remove the catalyst and saponified products from free fatty acids. Besides, water-containing waste oils/fats depress the catalyst activity. Thus, raw materials are limited to water-free and fatty acids-free oils/fats. For solving these problems, as a NEDO national project, R & D for the two-step supercritical methanol method (Saka-Dadan process) has been started in FY2003 for coming 3 years. Through the industry-university joint research, our research group has been working for establishment of the commercial process to produce high-quality biodiesel fuel (total glycerol content <0.24wt%) from a variety of oils/fats including crude and waste oils/fats with a high energy efficiency (>80%). In this paper, the achievements in this NEDO research project are introduced. Keywords: Biodiesel, Fatty Acid Methyl Esters, Esterification, Supercritical Methanol, NEDO Project 1. INTRODUCTION Biodiesel fuel (BDF) is one of the most promising bioenergies, which can be produced from oils/fats through transesterification. In Japan, about 3.0 million tons of oils/fats were consumed mainly as food products, of which about 2.6 million tons were imported from abroad in 2003 [1]. In such a heavy dependence of oils/fats resources, these wastes should be used efficiently as raw materials for BDF, instead of virgin ones. However, they sometimes include free fatty acids (FFA) and water, which have negative effects on yield of BDF. Especially in the alkali-catalyzed method, which is a current commercial process, FFA reacts with alkali catalyst to form saponified products, while water inhibits catalyst activity. Therefore, this conventional process cannot allow low-quality feedstocks containing FFA and/or water, such as waste oils/fats from the rendering plants called as dark oil in this study. On the other hand, a high-quality BDF must be prepared, which satisfies standard specifications such as EN 14214 and ASTM D 6751 to guarantee the quality of BDF and avoid any deterioration in the engine system for safe driving. At present, however, no national standards have been established in Japan, while a provisional specification exists only in Kyoto (Kyoto Standard), a pioneer municipal city for BDF utilization in Japan. With increasing concerns to BDF utilization, therefore, a working group has been organized by Ministry of Economy, Trade and Industry since January 2003 to discuss a national regulation for BDF/diesel blends, and it is planned to establish and issue a set of technical standards for neat BDF by the end of FY2005 [2]. In these situations, Kyoto University has been working to develop non-catalytic biodiesel production process by supercritical methanol treatment (>239 o C, >8.09MPa) since 1998 to produce high-quality BDF from a variety of oil/fat feedstocks including their wastes. Variables affecting the transesterification reaction of oils/fats were investigated followed by a proposal of the optimum conditions by supercritical methanol method to be 350 o C/20~50MPa and 9 min for transesterification of oils/fats [3-6]. Compared with the alkali-catalyzed method, this one-step supercritical methanol method (Saka process) has superiorities in terms of reaction time and purification step. In addition, the yield of BDF is high because of simultaneous methyl esterification of FFA to fatty acid methyl esters (FAME). These findings indicate that the supercritical methanol would provide a clue as to establishment of the efficient biodiesel production process. However, the one-step method requires a restrictive reaction condition compared to the conventional alkali- and acid-catalyzed methods. Therefore, further effort was made to develop an alternative method through the two-step preparation; hydrolysis of triglycerides in subcritical water and subsequent methyl esterification of fatty acids (FA) in supercritical methanol (Saka-Dadan process) [7]. The proposed reaction conditions of this method were 270 o C/7~20MPa for hydrolysis and methyl esterification. For such milder conditions, the common stainless steel can be applicable and resist enough for reaction vessel. Furthermore, the BDF product obtained by this method was low in total glycerol content which satisfies the specifications of BDF in EU and US standards. This process has been, thus, selected as one of the High Efficiency Bioenergy Conversion Projects by New Energy and Industrial Technology Development Organization (NEDO) for its practical use. In this paper, the NEDO national project is introduced on its current R & D status. 2. SCHEME OF THE NEDO PROJECT The NEDO project has been started since September 2003 with a total budget of about 840 million ( 6 million). Fig.1 shows the framework of the industry-university joint research team. In this project, Asahi Kasei Corp. is responsible for establishment of the commercial process to produce high-quality BDF (total glycerol content < 0.24wt%) through a high energy efficiency (>80%), while Kyoto University is mainly for fundamental study on the reactions in the two-step method and BDF evaluation, Asahi Engineering Co., Ltd. is for development of the automatic/safety control systems, while Asahi Research Center Co., Ltd. for establishment of the business model on BDF and Toyota Tsusho Corp. for market research on the oil/fat feedstocks and BDF, respectively. Corresponding author: saka@energy.kyoto-u.ac.jp This paper is also presented at the 14th European Biomass Conference held in Paris, France in October 2005. 1

NEDO Budget : 840 million Period : 2003.9-2006.3 Asahi Kasei Corp. Development of commercial process Kyoto University Fundamental study and BDF evaluation Asahi Engineering Co., Ltd. Automatic/safety control systems Asahi Research Center Co., Ltd. Establishment of business model Toyota Tsusho Corp. Survey on oil/fat feedstocks available for BDF Fig.1 Industry-university joint research groups as of FY2003 in the NEDO national project for high-quality BDF production by the two-step supercritical methanol method 3. RESEARCH ACHIEVEMENTS Two flow-type systems have been designed and developed based on scientific and technical knowledge at Kyoto University; one is the laboratory-scale equipment for the fundamental study on the reactions, while the other is the bench-scale plant with a product purification step, which is used to discuss a mass production technology. In addition, pre- and post-treatment steps for oil/fat feedstocks and waste water, respectively, were examined to establish the total process of the system. On the other hand, a trend survey on BDF including its production, utilization and policy was made in Japan as well as foreign countries. A market research on oil/fat feedstocks was also carried out to assure the feedstock suppliers. 3.1 Development of the practical total process (Asahi Kasei Corp.) As a preliminary study, economical efficiency of the two-step method was theoretically estimated. As a result, it was found to be comparable to that by the conventional alkali-catalyzed method. For the process design of the bench-scale plant, on the other hand, the fundamental reaction data obtained by Kyoto University was analyzed, mainly focusing on the mass transfer, phase equilibrium conditions and water distribution equilibrium. The benchscale plant with a product purification step (Fig.2) was then developed in Fuji City, Japan, in which the tubular and cylindrical reactors were employed. The capacity of this plant is about 50 L/day on feedstock basis. To realize a continuous operation safely, the emergency shutdown system with safety interlocks was also installed. Refined rapeseed oil as well as waste oils/fats collected by Toyota Tsusho Corp. were mainly used as a raw material to obtain basic data on the two-step reaction. 3.2 Fundamental study on hydrolysis and methyl esterification reactions (Kyoto University) Behaviors in hydrolysis and methyl esterification were studied in newly developed laboratory-scale system (Fig.3) for optimizing their reaction conditions. The effects of temperature, pressure and molar ratio on the hydrolysis in subcritical water was investigated by using refined rapese ed oil. It was, consequently, found that the conversion of rapeseed oil to FA can be made successfully at the conditions of 270~310 o C, 7~20MPa and 50~60 in the molar ratio of water to oil. In addition, the rate of FA formation was found to increase with the duration of the hydrolysis reaction, though it was very slow in early stage of the reaction. This phenomenon can be explained by assuming that FA acts as acid catalyst in the hydrolysis step [8]. For the methyl esterifcaiton reaction in supercritical methanol, on the other hand, the conversion to FAME could be made success fully at 270~290 o C, 7~20MPa and 10~20 in the molar ratio of methanol to FA. As a similar manner of the hydrolysis reaction, the m ethyl esterification was found to proceed by the autocatalytic mechanism. In this way, FA is a key compound in the two-step method. In case of the transesterification of triglycerides in supercritical methanol (one-step method), however, the reaction proceeded according to a typical pseudo-first-order reaction, since there is no FA in the reaction system. That is a main reason why the two-step method can realize milder reaction conditions than those of the one-step method. 2

Methanol (Methanol recovery) Distillation column Reactor 2 Water (return to water tank) Oils/fats Separator Biodiesel Water Reactor 1 Waste water (with glycerol) Fig. 2 Photograph and schematic diagram of bench-scale plant of the two-step supercritical methanol process in Fuji City, Japan Fig. 3 The laboratory-scale two-step supercritical methanol system at Kyoto University, Japan 3

3.3 Development of the automatic/safety control systems (Asahi Engineering Co., Ltd.) Under high temperature and high pressure conditions of the supercritical fluid, a secured operation of the system is an important concern. Therefore, automatic operation and safety control systems were designed and applied to the bench-scale plant. A concept of remote supervisory control system was organized through a case study, assuming the high temperature and high pressure conditions, especially to detect the first indication of the problem on pump operation. 3.4 Trend survey on BDF in Japan (Asahi Research Center Co., Ltd.) To establish the business model for BDF production, a trend survey was firstly carried out thorough a field interview for BDF manufacturers in Japan as well as foreign companies. As a result, it could be summarized that, in Japan, BDF is produced in small-scale mainly for domestic use. As main raw materials, waste edible oils/fats from household sector are collected, and simply filtrated to remove impurities. However, most facilities do not evaluate their BDF product routinely. They need national standard for BDF evaluation. Regarding to the economical aspects, BDF is competitive with conventional diesel fuel if the feedstock cost is null. In fact, however, the production cost in Japan is rather high. Almost BDF production facilities in Japan request financial supports on initial cost as well as tax incentive. In this situation, therefore, BDF is produced mainly for environmental protection and recycling of waste edible oils/fats. 3.5 Survey on waste oil/fat feedstocks and BDF utilization (Toyota Tsusho Corp.) To ensure enough supply of oil/fat feedstocks, eligible suppliers and their waste oil samples were investigated and collected, respe ctively. Waste oils/fats were, then, analyzed to deterimine iodine value, fatty acid composition, acid value, water content, peroxide va lue and so on, whose values are highly related with the quality of BDF [9]. However, since the market of waste oils/fats has not been well established yet, their qualities, characteristics and costs highly depend on the suppliers. 4. BIODIESEL EVALUATION Table 1 shows typical BDF properties produced from waste rapeseed oil and dark oil by using the bench-scale system. Since the waste rapeseed oil contains very small amount of FFA and water, it can be considered as a good raw material as virgin ones available even for the alkali-catalyst method. For rapeseed oil-derived BDF, actually, the total glycerol content became less than 0.24wt% after the two-step reaction, which satisfies the requirement in EN standards. Concomitantly, almost other fuel properties in Table 1 also fulfilled their specifications. In the two-step method, glycerol is removed after the hydrolysis so the backward reaction of FAME can be depressed in the methyl esterification step [7]. On the other hand, the dark oil, which is yielded from oil/fat rendering plant as a by-product containing large amount of FFA (>60wt%) with various contaminants, phospholipids and sulfurated lipids, cannot be converted into BDF at all by the alkali-catalyzed method. In case of the two-step method, however, the conversion was made successfully as shown in Table 1. This is a characteristic advantage of this two-step supercritical methanol method, compared with the conventional alkali-catalyzed method. The high-quality BDF can be, thus, achieved by this process from a variety of feedstocks. Table 1 BDF evaluation prepared by the two-step method using the bench-scale supercritical methanol system from waste rapeseed oil and dark oil Properties Unit EN 14214 Raw materials *a Rapeseed oil Dark oil Density g/ml 0.86~0.90 0.883 0.883 Viscosity *b mm 2 /s 3.5~5.0 4.70 4.41 Pour point o C - -7.5-2.5 Cloud point o C - -8-2 CFPP o C - -8-3 Flash point o C >120 172.5 160.5 Carbon residue *c wt% <0.3 0.04 0.04 Cetane number - >51 54 50 Ester content wt% >96.5 99.5 96.1 Total glycerol wt% <0.25 *d *d Water content wt% <0.05 0.04 0.03 Methanol content wt% <0.2-0.011 Sulfur mg/kg <10 <3 14 Oxidation stability hrs >6 >>6 *e 8.8 *e Acid value mg/g <0.5 0.32 0.29 Iodine value g/100g <120 99 107 Calorific value *f kj/g - 39.7 39.7 *a: purification step was adapted after esterification reaction, *b: measured at 40 o C, *c: on 10% fuel basis, *d: not detected, *e: antioxidant was added, *f: gross calorific value 4

However, only acid value was rather high and could not be in a range below 0.5 mg-koh/g specified in EN 14214. At present, therefore, FA removal or re-esterification step is required to satisfy the specification for acid value, and the results in Table 1 is the ones after purification treatment. As previously noted, since the methyl esterification proceeds mainly through the autocatalytic mechanism of FA, the rate of FAME formation becomes low when the reaction proceeds extensively. In addition, a backward reaction of FAME to FA exists due to the presence of water formed by the methyl esterification. For these reasons, acid value by the two-step method tends to be rather high so that the additional treatment process is now under examination through the NEDO project for practical use of the two-step method. 5. SUMMARY The NEDO national project has been started since September 2003 for practical use of the two-step supercritical methanol method, which allows a variety of oils/fats including their wastes as raw materials with relatively moderate reaction conditions. Through the project, a bench-scale and laboratory-scale plants were designed and developed in Asahi Kasei corp. and Kyoto University, respectively. The development of scale-up technology and fundamental study on the two-step reaction are, thus, working by the industry-university joint research. Although the problem on acid value remains unsolved, the obtained BDF can mostly satisfy the specification standards in EU and US. Therefore, the BDF production process by this two-step method can be expected for the commercialization as a result of the NEDO achievements. 6. ACKNOWLEDGEMENTS This work has been done as one of the NEDO High Efficiency Bioenergy Conversion Projects. In addition, the supercritical methanol methods (one-step and two-step methods) have been firstly developed through the Grant-in-Aid for Scientific Research (B)(2) (No. 13556058, 2001.4~2003.3) and the Kyoto University 21COE program Establishment of COE on Sustainable-Energy System granted from the Ministry of Education, Science, Sports and Cluture, Japan, for all of which the authors are highly acknowledged. 7. REFERENCES [1] Ministry of Agriculture, Forestry and Fisheries, Japan (2004) Wagakuni no Yushi-jijo (Present status of oils/fats in Japan), pp. 88-89 (in Japanese). [2] Ministry of Economy, Trade and Industry, Japan (2005) 8th meeting minutes of the working group for standardization of automobile fuel on 8 July, 2005 [3] Saka, S. and Kusdiana, D. (2001) Biodiesel fuel from rapeseed oil as prepared in supercritical methanol, Fuel, 80, pp. 225-231. [4] Kusdiana, D. and Saka, S. (2001) Kinetics of transesterification in rapeseed oil to biodiesel fuel as treated in supercritical methanol, Fuel, 80, pp. 693-698. [5] Kusdiana, D. and Saka, S. (2001) Methyl esterification of free fatty acids of rapeseed oil as treated in supercritical methanol, J. Chem. Eng. Jpn., 34, pp. 383-387. [6] Tabe, A., Kusdiana, D., Minami, E. and Saka, S. (2004) Kinetics in transesterification of rapeseed oil by supercritical methanol treatment, Proc. 2nd World Biomass Conf. Exhib., pp. 1553-1556. [7] Kusdiana, D. and Saka, S. (2004) Two-step preparation for catalyst-free biodiesel fuel production: Hydrolysis and methyl esterification, Appl. Biochem. Biotechnol., 115, pp. 781-791. [8] Minami, E. and Saka, S. (2005) Kinetics of hydrolysis and methyl esterification for biodiesel production in two-step supercritical methanol process, Fuel, 85, pp. 2479-2483. [9] Imahara, H., Minami, E., Hattori, M., Murakami, H., Matsui, N. and Saka, S. (2005) Survey on waste oil/fat feedstocks for biodiesel production, Proc. 14th Euro. Biomass Conf. Exhib., pp. 1427-1430. 5