A Continuous-flow Reactive Distillation Reactor for Biodiesel Preparation from Seed Oils

Transcription

1 An ASAE/CSAE Meeting Presentation Paper Number: A Continuous-flow Reactive Distillation Reactor for Biodiesel Preparation from Seed Oils Arvinder P. Singh, Joe C. Thompson, B. Brian He Biological & Agricultural Engineering Department, University of Idaho Moscow, ID ,USA Written for presentation at the 2004 ASAE/CSAE Annual International Meeting Sponsored by ASAE/CSAE Fairmont Chateau Laurier, The Westin, Government Centre Ottawa, Ontario, Canada 1-4 August, 2004 Abstract. In biodiesel preparation from vegetable oils and alcohol through transesterification process in the presence of a catalyst, excess alcohol, typically 100% more than the theoretical molar requirement, is used in existing batch and continuous-flow processes in order to drive the reversible transesterification reaction to a high enough conversion rate. The excess alcohol needs to be recovered in a separate process which involves additional operating and energy costs. In this study, a novel reactor system using reactive distillation (RD) technique was developed and studied for biodiesel preparation from yellow mustard seed oil. The main objective was to dramatically reduce the use of excess alcohol in the feeding steam, which reduces the cost in downstream alcohol recover processes, and meanwhile maintain a high alcohol-to-oil molar ratio inside of the RD reactor, which ensures the completion of the transesterification of seed oil to biodiesel. A lab scale sieve-tray RD reactor system was developed and used in this study. Process parameters were studied on the effect of reduced alcohol to oil ratio on the overall quality of biodiesel product and the efficiency of such an RD reactor. Product parameters such as methyl ester content, viscosity, total glycerol, and methanol content were analyzed as per ASTM methods. Preliminary results showed that process parameters of methanol-to-oil ratio of 4:1 (molar) and a column temperature of 65 C produced a biodiesel that met the ASTM standards for total glycerol and viscosity. Keywords. Biodiesel, Reactive Distillation, Transesterification, Seed Oils The authors are solely responsible for the content of this technical presentation. The technical presentation does not necessarily reflect the official position of the American Society of Agricultural Engineers (ASAE), and its printing and distribution does not constitute an endorsement of views which may be expressed. Technical presentations are not subject to the formal peer review process by ASAE editorial committees; therefore, they are not to be presented as refereed publications. Citation of this work should state that it is from an ASAE meeting paper. EXAMPLE: Author's Last Name, Initials Title of Presentation. ASAE Paper No. 04xxxx. St. Joseph, Mich.: ASA E. For information about securing permission to reprint or reproduce a technical presentation, please contact ASAE at hq@asae.org or (2950 Niles Road, St. Joseph, MI USA).

2 Introduction A CONTINUOUS-FLOW REACTIVE DISTILLATION REACTOR FOR BIODIESEL PREPARATION FROM SEED OILS Arvinder P. Singh, Joe C. Thompson, B. Brian He 1 It has necessitated the governments, research communities, and private organizations around the world to look for alternative and renewable sources of energy due to the depletion of petroleum reserves, increase in energy demands, unpredictability of fossil oil production, and increased concerns of rising greenhouse gas emissions. To date, many alternatives have been researched and demonstrated but only a few have been proven to be practically feasible in terms of availability, economics, public and environmental safety, and simplicity of use. One such possible alternative is biodiesel from vegetable oils, used at 100% or blended with diesel fuel for compression-ignition type engines. Several studies have showed that biodiesel is a better fuel than fossil-based diesel in terms of engine performance, emissions reduction, lubricity, and environmental benefits (Peterson et al., 1997; Canakci and VanGerpen, 2000). Biodiesel can be made from vegetables oils or animal fats though transesterification or alcoholysis, enzymatic or lipase conversion, and thermal cracking or pyrolysis. In pyrolysis method, fatty oils molecules were thermally or catalytically converted into hydrocarbons mainly alkanes and alkenes, which are further fractionated to produce biogasoline and biodiesel. The equipment and operating cost for pyrolysis is expensive. The most commonly used method is transesterification of vegetable oils or fats with methanol or ethanol in the presence of a catalyst. The reaction is shown in fig. 1. R2 OOC CH CH2 OOC R1 + 3 R4 OH Catalyst HO CH CH2 OH R 1COO R 4 + R 2COO R 4 CH2 OOC R3 CH 2 OH R 3 COO R 4 Seed oil Alcohol Glycerol Esters (triglycerides) Figure 1. Transesterification of seed oils to produce fatty acid esters Because the reaction is reversible, excess alcohol is used to the shift the equilibrium to the products side. The completion of the transesterification reaction involves multiple parameters including the molar ratio of oil-to-alcohol, catalysts, reaction temperature, reaction time, and free fatty acids and water content of oils or fats. The mechanism and kinetics of biodiesel production have been studied by many researchers (Noureddini, 1997; Darnoko et al., 2000; Freedman et al., 1986). These studies show that transesterification consists of a number of consecutive, reversible reactions. Triglycerides are first reduced to diglycerides. The diglycerides are subsequently reduced to mono-glycerides. Lastly, the mono-glycerides are reduced to fatty acid esters and glycerol. The order of the reaction changes with the reaction conditions. Alkali-catalyzed batch transesterification process is simple and, therefore, was often used in commercial processes (Ma and Hanna, 1999). The reaction mechanism and effects of process 1 Corresponding Author, Biological and Agricultural Engineering, University of Idahom,81A JML, Moscow, ID ; Phone: ; Fax: ; bhe@uidaho.edu 2

3 parameters have been widely studied. Most of the studies were published as patents. At the University of Idaho, Peterson et al. (1991) characterized a batch transesterification process of winter rapeseed oil. A 6:1 molar ratio of methanol to oil or 100% excess gave the best conversions in 60 min in a batch system. Ma et al. (1998a) reported the reaction proceeded very fast from one to 5 min then slowed down and reached the maximum conversion at about 15 min when a ratio of 6 moles of methanol per mole of oil and 0.3% NaOH were used. Freedman et al. (1984) transesterified peanut, cottonseed, and sunflower oils used the same methanol to oil ratio and 0.5% sodium methoxide as catalyst at 60ºC. A yield of about 80% on soybean and sunflower oils was observed after 1 min. Batch methods for biodiesel production are slow, tedious, labor intensive, and low in productivity. Continuous transesterification processes are preferred over batch processes in commercial production. The basic advantages of continuous-flow process are a greater productivity and a consistent product quality. Continuous transesterification of seed oils was studied as early as the 1940s (Trent, 1945; Allen et al., 1945). It involves transesterification, separation of the co-product glycerol, washing of the ester product, and the recovery/recycle of the excessive alcohol. Such a continuous process usually involves an elevated temperature (120 C to 160 C). Recently, Noureddini et al. (1998), Peterson et al. (1999c), and Darnoko and Cheryan (2000) have also investigated continuous transesterification processes using different feedstocks. At the University of Idaho, a continuous biodiesel process using ladder type retention reactor was demonstrated (Peterson et al., 2000). The process utilized rapeseed oil and ethanol input at 1:6 molar ratio. An optimization study for biodiesel production by sunflower oil transesterification conducted by Antolin et al. (2002) used three times the stoichiometric quantity of methanol (0.28% w/w of potassium hydroxide to oil), 70 C temperature and two washings, one with slightly acidic (phosphoric) water and the other with pure water. Almost all of the existing processes utilize excess alcohol, which must be recovered in additional separation process. This recovery process generally involves distillation, which increases energy consumption and process time. This research explores the applicability of homogenous reactive distillation (RD) technique for transesterification of seed oils for biodiesel preparation. Reactive distillation is a technique of simultaneous implementation of chemical reactions and distillation in a counter-current column. This process has less recycle streams and a reduced need for waste handling, which translates into lower investment and operating costs. In some applications particularly when reversible reaction equilibrium prevents high conversions, the RD technique can be employed to remove the reaction products from the reaction zone thus improves overall conversion rate and selectivity. In other applications, reactions are utilized to overcome the separation problems caused by azeotropes. Some of the representative applications of RD techniques are in the production of methyl tertiary butyl ether (MTBE), ethyl tertiary butyl ether (ETBE), tertiary amyl methyl ether (TAME), methyl acetate, ethyl acetate, and butyl acetate. The operation of RD process is quite complicated and its performance is influenced by several parameters including the size of reaction and separation zones, reflux ratio, and feed rate and tray location (Solokhin, 1996; Tuchlenski et al., 2001). A conceptual design of catalytic reactive distillation for fatty acid esterification was discussed by Omata et al. (2003). The RD process can be performed in either tray or packed columns, however, tray columns are recommended for homogenous systems because of the greater holdup and the associated longer residence time. For reactions that are not severely equilibrium-limited, the initial reaction rate is high, but declines abruptly when compositions and temperatures approaches equilibrium. For such cases a pre-reactor should be used upstream of the reactive column to handle a substantial part of the reaction duty, which greatly enhances the RD technique. 3

4 There is a large difference in the boiling points of alcohol (either methanol or ethanol) and the seed oils or their fatty acid esters involved in the transesterification reaction (table 1). During the operation, most of the methanol would be in a vapor phase while the conversion to biodiesel would happen in a liquid phase. Table1. Boiling points (at 1 atm) of chemical components involved during the transesterification reaction ( a Goodrum, 2002; b Merck Index, 9th ed.) Component Rapeseed methyl ester Canola methyl ester Methanol Glycerol Bioling Point ( C) a a 64.7 b b Figure 2 illustrates the general setup of RD technique applied to biodiesel preparation in a trayed-column. The feed stream of oil and methanol/catalyst was introduced at the upper tray of the column and, combining with the condensed alcohol vapor, flows downward; while the methanol vapors from the reboiler flask rises upwards causing a counter-current gas-liquid contact. Each tray in the column forms a reactive zone, or a mini-reactor. Each mini-reactor contains alcohol which is much more concentrate than that in the feeding stream, forming a high alcohol to oil ratio and resulting in a more uniform and higher reaction rate. Condenser Feed Alcohol Vapors Partially reacted mixture Reboiler Product Figure 2. Reactive distillation column operated counter-currently. The ultimate goal of this research is to explore a technically and economically sound reactor technology for biodiesel production, which applies the reactive distillation technique. The objectives of this study were to (1) construct a lab-scale continuous-flow reactive distillation process system, and (2) determine the effect of oil-to-alcohol ratio on the fatty ester yield by targeting at a molar ratio close to the theoretical ratio of 3:1. 4

5 Materials and Methods The crude canola oil used in this research was obtained from the oil seed processing plant at the Department of Biological and Agricultural Engineering of the University of Idaho. Analytical grade methanol and potassium hydroxide were of obtained from J.T. Baker (Phillipsburg, NJ). GPO-trinder Reagent and reference standards such as triolien, diolien, methyl olieate and glycerol were purchased from Sigma-Aldrich Co. (St.Louis, MO). Table 2. Fatty Acid Profile of Canola Oil used in this research Fatty Acid Composition (% wt.) Palmitic (16:0) 3.9 Stearic (18:0) 2.1 Oleic (18:1) 59.3 Linoleic (18:2) 18.4 Linolenic (18:3) 7.8 Eicosic (20:1) 2.1 Erucic (22:1) 4.4 Equipment and Experimental Setup A laboratory-scale continuous-flow RD system (fig. 3) was developed and tested on overall process parameters. The central system component is a vacuum jacketed, Oldershaw perforated plate distilling column (Chemglass, Vineland, NJ). This 10-plate column has an inner diameter of 23 mm, a weir height of 4 mm, and a distance of 25 mm between plates. Figure 3. Experimental setup of the laboratory-scale continuous-flow RD reactor system. 5

6 A short 3 ml in-line static mixer (Cole-Parmer, IL) was used as a feed mixer and a pre-reactor prior to the RD column. The lower end of the column was fitted to a 500 ml three-neck roundbottom flask was used as a reboiler. A water-cooled condenser was fitted to the top of the column to recover alcohol, which is combined with reactants and pumped back into the column. The product mixture was withdrawn from the reboiler and send to the separation column. Biodiesel/glycerol separation was carried out by gravity in a continuous decanter (φ mm) with an adjustable feed-in point. The input and output streams of methanol/koh solution, oil, and product mixture were handled simultaneously with three Masterflex peristaltic pumps (Cole- Parmer, IL) which were calibrated and adjusted to achieve the desired flow rates ranging from 0.5ml/min to 8.0ml/min. Temperature monitoring and feedback controls were accomplished with Fuji PXR3 and PXR4 PID/feedback controllers (distributed by TTI Inc., VT). Experimental Procedures Two sets of preliminary trials with and without a pre-reactor were conducted with the RD setup as shown in fig. 3. The main process parameters examined in this study were: oil-to-alcohol ratio, flow rates, and reaction time and temperature. In preparation for each trial, stock alcoholic KOH was prepared on a stirring plate at a ratio that corresponded to 1% KOH w/w of oil for each given methanol-to-oil molar ratio, and placed in a holding reservoir next to the RD column. Likewise canola oil was held in a separate heated reservoir maintained at 50 C. These reactants were fed through separate calibrated peristaltic pumps into the pre-reactor or directly into the column where the reaction began. The methanol-to-oil molar ratios used were 3.0, 3.5, 4.0 and 4.5. These various ratios were achieved by adjusting the alcohol/catalyst flow rate relative to the flow rate of the oil. For this column, residence (reaction) time can be related to overall flow rate. Additionally a suitable flow rate had to be maintained to avoid operational problems such as column flooding and weeping. From several trials, it was found that an overall flow rate of 5-6 ml/min with the column temperature at 65 C provided residence time of about 5 min without any significant operational difficulties. The column temperature was maintained by controlling the reboiler heat input. Temperatures above 65 C caused excessive entrainment and a reduction in methanol concentrations in the liquid phase. Lower temperatures caused significant reduction in the methanol vapor flow which directly affected the ester conversion. Excess methanol vapors were condensed and recycled back to the column while the product mixture was continuously drawn from the bottom flask. It was separated into biodiesel and glycerol layer by the continuous gravity decanter with feed-point at the middle of the column. Temperatures were monitored and controlled at various strategic locations, as shown in fig. 4. Recycle Methanol T4 T5 T2 Biodiesel T1 Oil Feed Methanol & KOH Feed mixer/ Pre-reactor Pre- RD Column T3 Decanter De Glycerol Gly Figure 4. Schematic diagram of the RD reactor system 6

7 Actual flow rates and temperature profiles used for each run, with varying methanol-to-oil ratio, were averaged and are shown in table 3. Trials Table 3. Process parameters used in different trials Molar Ratio 1 Q 1 oil Q methanol 2 T1 2 T2 2 T3 2 T4 2 T5 1) With Pre-reactor Run 1 3.0: Run 2 3.5: Run 3 4.0: Run 4 4.5: ) Without Pre-reactor Run 5 3.0: N/A Run 6 3.5: N/A Run 7 4.0: N/A Run 8 4.5: N/A 1 Qoil & Q methanol are Oil flow rate and methanol-catalyst flow rate, respectively, in ml/min. 2 T1, T2, T3, T4 & T5 are feed oil temperature, Column operating temperature, Reboiler temperature, column feed inlet temperature and pre-reactor feed inlet temperature, respectively, in degree Celsius. Product Characterization The product obtained from each run was analyzed for the contents of methyl esters, methanol, total glycerol, and the product viscosity and density. An HPLC (HP 1090 series) with ChemStation software and an evaporative light scattering detector (Altech2000; Deerfield, IL) was used to analyze the total methyl esters. The column was a Luna 5µ CN 100A, mm I.D (Phenomonex, Torrance, CA). The method, developed by Bruns (1989) was slightly modified for our purposes. It is a linear gradient method with mobile phase steps: 99.8% iso-octane + 0.2% 2-propanol initial 5 min hold up, 95% octane+5% 2-propanol in 15 min and held till 20 min, then finally 90% octane+10% 2-propanol in 30 min. Twenty milligram samples were diluted with a 10 ml solution of 96% octane+4% 2-propanol for a working sample dilution of 2mg/ml. Sample injection size was 20µl and both temperatures of the column and detector drift tube were maintained at 40ºC. Mobile phase flow rate was 1 ml/min and nebulizer gas flow rate was set at 1.5 l/min. The Greenhill method issued by National Biodiesel Accreditation Commission, BQP 02(03), was used to measured total glycerol content in biodiesel samples. It is a spectroscopic determination and was set up as an alternative to ASTM D 6584, which is specified under the standard specification for biodiesel fuel (ASTM D 6751). The specific gravity was measured using a 20-ml Bingham Pycnometer (Fischer Scientific, PA) at 20ºC under the guidelines of ASTM method D Viscosity measurements were made at 40ºC using a #100 Cannon-Fenske viscometer (Fischer Scientific, PA) set in a Koehler Model K oil bath (Koehler, Bohemia, NY). The methanol content in biodiesel samples was determined by the difference of a sample before and after heating to 70ºC for 30 min to drive off the alcohol. 7

8 Results and Discussions The process sensitive characteristics, such as total methyl esters, total glycerin, methanol, and viscosity, of samples obtained under different experimental runs were cited in table 4. Trials Table 4. The effect of methanol-to-oil ratio on the product parameters. Molar Ratio Methyl esters (%wt) Total Glycerol (% wt) Methanol (%wt) Viscosity (cst) Specific gravity 1) With Pre-reactor Run 1 3.0: Run 2 3.5: Run 3 4.0: Run 4 4.5: ) Without Pre-reactor Run 5 3.0: Run 6 3.5: Run 7 4.0: Run 8 4.5: As expected, the setup with a pre-reactor gave better reaction yield compared to corresponding runs without a pre-reactor. The methyl ester content increase with molar ratio justifies the need for excess alcohol to drive the reaction to a higher yield (fig. 5). Since, the reaction rates are quite high at the beginning, a major part of the reaction conversion can be achieved in the inline static mixer. The use of a pre-reactor also reduces the size of reactive zone and ultimately the size of the column and the cost of production. 100 Methyl Esters Content Methyl Esters (% wt) Methanol-to-Oil ratio without pre-reactor with pre-reactor Figure 5. Effect of methanol-to-oil ratio on methyl esters content in product. 8

9 A significant amount (approx. 10%) of unreacted methanol flowed out of the column with the products. About 30% of this methanol ended up in the biodiesel layer while the remainder was found in the glycerol layer. Figure 6 shows the trend followed by methanol content in the biodiesel layer. Biodiesel obtained without the pre-reactor had higher methanol content. This was most likely due to incomplete reaction due to less residence time. At a 4:1 molar ratio, methanol content in the product was found to be the lowest for both experimental setups under the tested conditions. The methanol content rose if the ratio was increased to 4.5:1 due to a greater excess of methanol and its incomplete vaporization from the reboiler. 6 Methanol Content Methanol (% wt) Methanol-to-Oil ratio without pre-reator with pre-reactor Figure 6. Effect of methanol-to-oil ratio on methanol content in product. There is a correlation between the viscosity and the amount of unreacted glycerides present in the biodiesel. It can be seen in table 3 that as the molar ratio increased, the percent ester increased while the unreacted glycerides correspondingly decreased. This is reflected in fig. 7 whcih illustrates a decrease in viscosity with a higher molar ratio. Additionally, samples obtained with the pre-reactor setup have higher ester percentages and hence lower viscosity as compared to corresponding runs without the pre-reactor. Kinematic 40C, cst Viscosity Profile Methanol-to-Oil ratio with pre-reactor without pre-reactor Figure 7. Effect of methanol-to-oil ratio on product viscosity. 9

10 Summary The reactive distillation process as it was set up and described in this paper, has been found to be feasible for the continuous production of biodiesel from seed oils. The original objective to make the process more efficient by reducing the alcohol-to-oil molar ratio was realized. A 66% reduction to the industrial standard of 6:1 was achieved with good results. From the data collected it can be concluded that the operating the RD system at 65ºC with a 4:1 molar ratio and with a pre-reactor was near the optimum point for producing biodiesel from among the parameters examined in this study. The biodiesel obtained under these conditions met the national standards (ASTM D 6751) for total glycerol and viscosity. Testing with this system will continue to in an attempt to reduce the energy and material inputs to improve the efficiency further while maintaining fuel quality. Acknowledgments This study was financially supported by the National Institute of Advanced Transportation Technology (NIATT) of the University of Idaho, and Idaho Rapeseed/Canola Commission. The authors would express appreciation to Andrej Paszczynski and David Christian for their assistance in the development of HPLC method for biodiesel product analysis. References Allen, H.D., G. Rock, and A. Kline Process for treating fats and fatty oils. U.S. Patent No Antolin, G., F.V. Tinaut, Y. Briceno, V. Castano, C. Perez, and A.I. Ramirez Optimisation of biodiesel production by sunflower oil transesterification, Bioresource Technology. 83: Bruns, A., H. Waldhoff, and W. Winkle Applications of HPLC with evaporative light scattering detection in fat and carbohydrate chemistry. Chromatographia 27(7/8): Canakci, M., and Jon VanGerpen The Effect of Yellow Grease Methyl Ester on Engine Performance and Emisssions. Final Report: Recycling and Reuse Technology Transfer Center. Publication # Darnoko, D., and Munir Cheryan Kinetics of Palm Oil transesterification in a batch reactor, JAOCS 77(12): Freedman, B., E.H. Pryde, and T.L. Mounts Variable affecting the yields of fatty esters from transesterified vegetable oils. JAOCS 61: Freedman, B., R.O. Butterfield, and E.H. Pryde Transesterification kinetics of soybean oil. JAOCS:63: Goodrum, J.W Volatility and boiling points of biodiesel from vegetable oils and tallow. Biomass and Bioenergy 22: Jackson, M.A. and J.W. King Methanolysis of seed oil in flowing supercritical carbon dioxide. JAOCS 61:

11 Ma. F., and M.A. Hanna Biodiesel production: a review. Bioresource Technology 70:1-15. Ma. F., L.D. Clements, and M.A. Hanna. 1998a. The effects of catalyst, free fatty acids and water on transesterification of beef tallow. Trans. ASAE 41: NBAC Designationation: BQP-02 (03). Standard Test Method for Free and Total Glycerol in Biodiesel (Greenhill Method). Noureddini, H., D. Harkey, and V. Medikonduru Continuous process for the conversion of vegetable oils into methyl esters of fatty acids. J. Am. Oil Chem. Soc. 75(12): Noureddini, H., and D. Zhn Kinetics of transesterification of Soyabean Oil, Applied Eng. in Agric. 74: Omata, F., A.C. Dimian, and A. Bliek Fatty acid esterification by reactive distillation. Chemical Engineering Science 58: Peterson, C. L., D.L. Reece, B. L. Hammond, J. Thompson, and S. M. Beck Processing, Characterization, and Performance of Eight Fuels from Lipids. Applied Eng. in Agric. 13(1): Peterson, C.L., J.L. Cook, J.C. Thompson, and J.S. Taberski Continuous Flow Biodiesel Production. Applied Eng. in Agric. 18(1): Peterson, C.L., M. Feldman, R. Korus, and D.L. Auld.1991.Batch type transesterification process for winter rape oil. Applied Eng. In Agric. 7(6): Solokhin, A.V., and S.A. Blagov Reactive-Distillation is an advanced technique of reaction process operation. Chemical Engineering Science 51(11): Tanaka, Y., A. Okabe, and S. Ando Method for the preparation of a lower akyl ester of fatty acid. U.S. Patent No Trent, W. R Process of treating fatty glycerides. U.S. Patent No Tuchlenski, A., A. Beckmann, D. Reusch, R. Dussel, U.Weidlich, and R.Janowsky Reactive Distillation industrial applications, process design & Scale-up. Chemical Engineering Science 56: