(659e) Heat Generation and Transfer Analysis of a Pilot Scale Reactor for the Production of Epoxidized Sucrose Soyate

Increased interest in epoxidized sucrose soyate (ESS) for applications testing required scaling up the production of ESS from the 300 g capability in our lab to 10 kg batches. The epoxidation process is exothermic and the removal of heat produced becomes an increasingly important consideration as the scale of reaction increases. Therefore, the heat flux from the reaction mixture must be sufficient to maintain the reaction temperature within the optimal epoxidation level. An analysis of the energy balance and heat transfer parameters is critical in managing the exothermic process during scale up while reducing reaction time and side product formation. The thermal and kinetic parameters of epoxidation have been characterized for a variety of vegetable oils at scales less than 100 g and with different types of reactor, but the parameters for sucrose soyate (sucrose ester of fatty acid from soybean oil) at a 10 kg pilot scale have not been characterized.

In an exothermic reaction like the epoxidation process, the simple expression for thermal energy balance using the 1^st law of thermodynamics in a batch reactor and considering the heat lose through evaporation is negligible, the expression is;

      Stored energy (rate of internal energy change) = generated energy – exchanged energy  

Where m_r is total mass of the reaction solution, c_r is the specific heat capacity of the total solution,  is the rate of change of the reactor temperature, U is the overall heat coefficient, A is the heat exchange surface area, T_r – T_f is the thermal gradient (DT) between reaction mixture and jacketed fluid. ΔE_r is the specific molar energy change due to reaction, and RV_r is the amount of moles converted in the reactor per unit time.

The optimum reaction temperature has been shown to be 55 – 65^oC in many studies where the average of 60^oC was used as the target reaction temperature. Therefore, the goal of this study was to analyze heat transfer and generation phenomena in a 45 L stainless steam-jacketed batch reactor that will ensure a constant reaction temperature at 60^oC for a total reaction time of 5.5 h; that is  = 0 and  

Heat generation in an epoxidation process depends on hydrogen peroxide (H₂O₂) addition rate. The heat generation rate was calculated at a fixed H₂O₂ addition rate of 0.75 g min^-1 mol^-1 unsaturated over 3 h addition time and three variable rates; 0.50, 0.75, and 1.0 g min^-1 mol^-1 unsaturated over 1 h addition time each, then reaction was continued for another 2.5 h. The rate constant (K) for epoxidation and ΔE_r were obtained from previous studies. The molar concentration of double bonds and peraceitc acid was determined from a small scale (50 g) experiment that analyzed the extend of epoxidation at every 15 min for the first 3 h. The intensity of heat transfer depends on the reactor design, stirring rate, and process conditions. Some of the reactor design parameters (size, jacketed area, type of material and thickness) could not be changed because the reactor was a steam-jacketed kettle that was redesigned and insulated as a reactor for the pilot scale production of ESS. The overall heat transfer coefficient (U) for the reactor was calculated when there was no heat generation; that is with all reactants except H₂O₂. Therefore, the jacketed temperature in the heat transfer equation was easily adjusted to obtain the required thermal gradient (DT).

Four batches of ESS would be produced at the different H₂O₂ rates (two each for the fixed and variable rates) and the experimental DT between reaction mixture and jacketed fluid temperatures would be compared with the theoretical DT for each time interval. The DT for the variable rate is expected to be smaller and closer to the theoretical value because our previous study on a 300 g scale showed that exothermic activities increased sharply within the first 20-30 min of H₂O₂ addition, and then starts dropping gradually even as H₂O₂ addition continues. Therefore, lowering H₂O₂ addition rate within the first 1 h of addition would be the best option.