Mixing Review

MIXING EFFICIECNY OF DIFFERENT AGITATOR TYPES: A SIMULATION Introduction In transesterifcation, it has been shown that the immiscibility of the oil and alcohol is a vital factor even at optimal elevated temperature and concentration and various studies have established the desirability to achieve 'immediate' and thorough establishment of an oil-alcohol inter-phase before the alkali is consumed and the reaction is brought to a stop. The relevance of fluid agitation is presented in Ma and Hanna's 1999 on the effect of mixing on transesterification of beef tallow. Studies [1,2] established that reaction rate was a function of mixing intensity for Reynold's number greater than 10,000, supporting also the mass transfered control theory during the initial stage of transesterifcation. transesterification kinetic model [3] predicted better conversion with more mixing intensity. biodiesel was produced from fryer grease [4] using a mixed methanol/ ethanol system at 600 rpm and [5] investigated effect of velocity for the transesterifcation of sunflower oil in a static mix and results obtained compared to those of mechanical agitation at 600 rpm. In a survey [6], eleven works highlighted agitation speed of 60-700 rpm for edible oil with alkali catalyst between 0.2-1.59% and alcohol-oil ratio of 3:1-6:1 for optimal transesterifcation. Here, drop size distributions were found to become narrower and shift to smaller sizes with increasing agitation speed as well as with the progress of the methanolysis reaction at a constant agitation speed. knowledge of mass transfer processes in transesterification is still far from complete and experiments at lower stirring speeds showed a reaction rate dependency on the stirring rate, indicating that the reaction rate is also mass transfer limited [7]. At higher stirring rates it was found that one phase was finely dispersed into droplets. The relevance of mixing is further emphasized in studies conducted in modeling the distribution in a liquid-liquid system, the dispersion pattern in the continuous and dispersed phases [8]. phase distribution of alcohol, glycerol and catalyst in the transesterification of soybean oil commencing as a two-phases reaction with an upper methanol phase, in which the catalyst is dissolved, and a lower vegetable oil phase was described [9]. Stirring initiates the reaction, which transforms to another two-phase system comprising an ester-rich phase and a glycerol-rich phase. When stirring is stopped, the glycerol-rich phase settles to the bottom. Optimization study of integrated biodiesel production set the impeller speed at 600 rpm to avoid mass transfer limitations on the process [10]. Dias (2008) simultaneously compared the catalyst performance of waste frying oil, sunflower and soybean refined oil using KOH, NaOH and CH3ONa as catalysts; optimum conditions were (i) 0.6 (wt%) CH3ONa for both virgin oils; (ii) 0.6 (wt%) NaOH for sunflower oil and 0.8 (wt%) for soybean oil; and (iii) 0.8 (wt%) using both catalysts for waste frying oils. Production using virgin oils results in higher yields (reaching 97%) as compared to waste frying oils (reaching 92%). From the above laboratory scale data, agitator type and orientation are seldom mentioned, relevant for industrial scale-up and application. According [11] the problem of scale-up can also be viewed as one of (model) uncertainty. The data available from laboratory studies do not quite extrapolate to the production level. Thus, when the strategies developed in the laboratory are used at the production level, they do carry a fair amount of uncertainty. According to Jakobsen [12], among the number of purposes for agitating fluids is to optimize blending of two immiscible liquids. There are a number of way of achieving this purpose. I will limit my highlight to mechanical mixing as it is the preferred and most common means of achieving mass distribution during transesterification both at bench scale and industrial operation. Mixing Design The investigation into the role of alcohol type & ratio, catalyst concentration, temperature and intensity of agitation as it affect the transesterification of fats and oil are well documented. The bench scale application of the data generated provides good start off point for Industrial operation. Most of these acid/ alkali-based transesterifcation process still rely on the original recipe of mixing excess alcohol with oil in the presence of a suitable alkali catalyst between 3:1 – 24:1 ratio for optimum and

equilibrium conversion. Although the production above the 6:1 alcohol to oil ratio has been found not to improve yield significantly for oils with FFS less than 2% Insight has been provided by studies into the various of process parameters. The conceptualization [13] of the ratio of alcohol, alcohol type and catalyst type and concentration formed the basis of subsequent transesterifcation works [14]. Even as later evidence have shown that the more heterogeneous catalyst provide faster and easier reaction, easiness of product separation, transportation and mobilization alkali-based transesterifcation remain a focus for large industrial transesterifcation process. However, the studies at improving this process has relied mainly on kinetic investigation of the components interaction irrespective of the agents involved. As noted rightly that no matter the type of catalyst is selected an intense mixing of the two agents is required in order to break the alcohol phase into small drops [5]. [2] study using mixing intensity of 150, 300 600 and 1240 rpm and and [15] In order to arrive at an optimum mixer design a detailed understanding of the mixing mechanisms involved and their importance in achieving the process result is required. These mixing mechanisms are described as either convection, macro-mixing, laminar shear and micro-mixing [16] where these mechanisms are determined by number of impellers, impeller positions, impeller diameter/ vessel diameter ratio (D/T), bottom clearance (C/T ratio), bottom shape, placement of baffles, mixing intensity (mild, moderate, vigorous, violent). The uniqueness in considering factors beyond the kinetic of chemical reaction is emphasized when chemically balanced reaction do not behave optimally. Most studies based on the use of the Rushton Impeller show that the power number is sensitive to the details of impeller geometry, and in particular to the blade thickness, but is independent of the impeller diameter to tank diameter ratio (D/T) whereas for pitched blade impeller the power number is independent of blade thickness, but dependent on the impeller to tank diameter ratio. This is exactly the opposite result to that observed for the Rushton turbine [17]. Chapple et al 2002 compared mixing effect of the geometric details for Rushton Turbine (RT) and Pitched Blade Turbine (PBT) using TriEthylene Glycol (4.1 X 10^-2 Pa s), Bayol (2.6 X 10^-3 Pa s) and Water (1.0 x 10^-3 Pa s). For RT, the effect of disk thickness is significant for 500 2x10^4 and concluded that Npft is not sensitive to blade thickness for the PBT impeller. PBT power draw is much less sensitive to details of the impeller geometry than the RT, it is much more sensitive to interactions with the tank walls. For both impellers, the importance of geometry decreases as the Reynolds number drops into the transitional regime and viscous forces come into play. Interactions between the PBT and the tank walls result in changes in the velocity field in the tank, found a decrease in Np with increasing D=T for the PBT. Hence effect of D=T on power number decreases as the Re is reduced to the low transitional range. The above are analogous with pipe flow and friction factors, that the effect of geometry on power number (and friction factor) persists at lower Reynolds numbers, but is less significant as the Reynolds number drops. It has been extensively demonstrated experimentally that the final product distribution in complex reactive systems is greatly affected by the fluid dynamics of the reactors (Baldyga and Bourne, 1990). Hence, variables, such as impeller type and position, feed location, agitation intensity, and feed time of the limiting reactant have been shown to significantly alter the final concentrations of the reaction products (Bourne and Yu, 1994).

Baldyga J., and J. R. Bourne, “The Effect of Micromixing on Parallel Reactions,” Chem. Eng. Sci. 45, 907 (1990).

SIMULATION OF REACTION [2] and [15] studies were adopted for this simulation.

Scale drawings of the Pitched Blade Turbine (PBT) and the Rushton Turbine (RT).

References 1. Fangrui Maa, Milford A. Hanna. BIODIESEL production: a review. Bioresource Technology 70 (1999) 1-15 2. H. Noureddini and D. Zhu. Kinetics of Transesterification of Soybean Oil. JAOCS 74, 1457– 1463 (1997). 3. Matthew Slinn and Kevin Kendall. Developing the reaction kinetics for a BIODIESEL reactor. Bioresource Technology 100 (2009) 2324–2327 4. Titipong Issariyakul, Mangesh G. Kulkarni, Ajay K. Dalai, Narendra N. Bakhshi. Production of biodiesel from waste fryer grease using mixed methanol/ethanol system. Fuel Processing Technology. Volume 88, Issue 5, May 2007, Pages 429-436 5. Dario Frascari, Michele Zuccaro, Davide Pinelli, and Alessandro Paglianti. A Pilot-Scale Study of Alkali-Catalyzed Sunflower Oil Transesterification with Static Mixing and with Mechanical Agitation. Energy & Fuels 2008, 22, 1493–1501 6. Olivera S. Stamenkovic´, Zoran B. Todorovic´, Miodrag L. Lazic´, Vlada B. Veljkovic´, Dejan U. Skala. Kinetics of sunflower oil methanolysis at low temperatures. Bioresource Technology 99 (2008) 1131–1140 7. Devender S. Negi Æ Felix Sobotka Æ Tobias Kimmel Æ Gu¨nter Wozny Æ Reinhard. Schoma¨cker. Glycerolysis of Fatty Acid Methyl Esters: 1. Investigations in a Batch Reactor. J Amer Oil Chem Soc (2007) 84:83–90 8. Fredrik J. E. Svensson and Anders Rasmuson. PIV measurements in a liquid–liquid system at volume percentages up to 10% dispersed phase. Experiments in Fluids: Experimental Methods and their Applications to Fluid Flow. Volume 41, Number 6 / December, 2006 9. Weiyang Zhou and D.G.B. Boocock. Phase Distributions of Alcohol, Glycerol, and Catalyst in the Transesterification of Soybean Oil. JAOCS, Vol. 83, no. 12 (2006) 10. Gemma Vicente, Mercedes Martı´nez, Jose´ Aracil. Optimisation of integrated BIODIESEL production. Part II: A study of the material balanc. Bioresource Technology 98 (2007)

11. B. Srinivasan, D. Bonvin, E. Visser, S. Palanki. Dynamic Optimization in the Batch Chemical Industry, Role of measurements in handling uncertainty. Computers and Chemical Engineering 27 (2003) 27–44. Computers & Chemical Engineering, Volume 27, Issue 5, 15 May 2003, Page 761, 12. H.A. Jakobsen, Chemical Reactor Modeling. Handbook of Industrial Mixing. Chapter 7 Agitation and Fluid Mixing Technology. Springer-Verlag Berlin Heidelberg 2008 13. Freedman 14. Darnoko, D., and M. Cheryan, Kinetics of Palm Oil TRANSESTERIFICATION in a Batch Reactor, J. Am. Oil Chem. Soc. 77:1263–1267 (2000). 15. Gwi-TAEK JEONG AND DON-HEE PARK. Batch (One- and Two-Stage) Production of BIODIESEL Fuel From Rapeseed Oil. Applied Biochemistry and Biotechnology, Vol. 129-132, 2006

16. www.mixingslutions.com 17. D. CHAPPLE, S. M. KRESTA, A. WALL and A. AFACAN. THE EFFECT OF IMPELLER AND TANK GEOMETRY ON POWER NUMBER FOR A PITCHED BLADE TURBINE. Institution of Chemical Engineers Trans IChemE, Vol 80, Part A, May 2002 18. Joana M. Dias, Maria C.M. Alvim-Ferraz, Manuel F. Almeida. Comparison of the performance of different homogeneous alkali catalysts during transesterification of waste and virgin oils and evaluation of biodiesel quality. Fuel 87 (2008) 3572–3578