(612d) Thermal Cracking of Frying Oil: A Proposal for a Kinetic Mechanism Based on Groups of Compounds

The growing concern for the environment has made efforts in search of clean technologies and renewable fuels. In this way, thermal cracking of organic wastes is an alternative to reduce the volume of waste generated, while at the same time, being able to generate energy through it.

Thermal cracking can process, as raw material, a wide class of wastes of difficult treatment, like plastics, waste oils and animal fats, municipal solid waste, certain components of electronics, among others. The thermal cracking process always generates three streams of products: liquid, gas and solid fractions. The yields of each one varies in function of temperature and residence time. The liquid fraction decreases with increasing temperature due to the cracking reactions, which result in the formation of organic products of low molecular weight. At high temperatures, these products are converted into CO2, H2, CO, CH4, C2H6 and C3H8, etc., and hence the gas production increases with temperature. In the thermal cracking reaction of fats, the breakdown of triglyceride molecules leads to the formation of lower molecular weight compounds. The liquid fraction produced in the cracking process is called bio-oil.

This bio-oil has a complex composition, with different chemical classes of hydrocarbons. It is the result of the degradation of biomass by heat, and it has specific properties and a high calorific value. However, to be used as biofuel, it should be fractionated, in order to separate fractions of compounds for fuels in the range of petrol and diesel. When performed at temperatures in the range of 500 °C, the reaction produces a bio-oil with similar properties to those of fossil fuels. The kinetic model proposed for this type of reaction needs to take into consideration the type of raw material used, operational conditions, such as temperature, residence time and reactor configuration, with the model being function of these properties and operating conditions. A suitable model which allows the prediction of these fractions is useful for equipment building design and operational management of the plant. Models for these reactions involve many variables and are difficult to determine. One strategy that can be applied to these studies aiming to reduce these variables is the lumping (grouping of chemical species), which consists of grouping variables according to a linear or nonlinear function. Generally, each lump describes a hydrocarbon fraction resulting from the thermal cracking and its particular kinetic mechanism. These models are good for any study of the thermal cracking reaction, as they provide information about the reactions and their corresponding parameters, process efficiency and product generation. The complexity involved in the construction of these models is related to the number of parameters to be experimentally described. The type of reaction or various reactions involved in the process approach will determine, through experimental verification of the applicability, the validity of the model. Simple models with a small number of lumps (3 to 5) facilitate the numerical analysis for the determination of relevant parameters of the reaction. The numerical and experimental data that has been adjusted in this work are based on the mass fraction of the products of thermal cracking, using mass balance and chromatographic analysis. The regression method used in this study was the Complex method, which aims to direct search of the optimum value (maximum or minimum) of an objective function with n variables. To check the numerical and experimental data, a parameter used to qualify the regression is the value of the objective function, which is defined as the sum of the square of errors among experimental values and simulated values. The difference between then should be the lowest possible.

This work intends to propose kinetic models for the thermal cracking reaction of frying oil. The pyrolytic reactor used to obtain experimental data consists of a cylindrical tube with a rotating screw inside and it is operated in a continuous mode at an isothermal and a steady state condition. For the purpose of kinetic parameters, the reactor was considered as a PFR (plug flow reactor). The raw material used for the thermal cracking was frying oil. Based on results presented in the literature, three work temperatures were defined (500, 525 and 550 °C) and for each temperature, 10 different biomass feed flows were applied, resulting in different spatial times. Flows operated in each temperature were in the range of 120 g/h and 500 g/h, for the lowest and the highest flow rate, respectively. So, ten experiments were performed at each temperature, totaling thirty experiments, conducted in triplicate under the same conditions of the original experiments, totaling ninety experiments. The experimental variables were the reaction temperature, input flow of raw material in the reactor and yields of converted products. To close the mass balance, it was necessary to quantify the bio-gas generated in the process. This measurement was based on volumetric flow rate and the chemical composition of bio-gas. In order to get these compositions, gas chromatography assays were performed. Quantification of coke generated in the cracking is the result of the difference between the initial biomass, bio-oil and bio-gas yields. So, five groups of compounds based mechanisms (with four lumps) have been proposed. All mechanisms have proved satisfactory for the physical representation of the process of thermal cracking, especially models 2 and 3. The objective function value and the graphics between the calculated and experimental data were compared, and based on this information the best models for representation of the process were set. The comparison between the five proposed mechanisms identified that the mechanism 2 with five parameters was the one, which best represented the reaction of thermal cracking of the frying oil, once it produced the smallest objective function value, and had the closest relation between the experimental and numerical values, representing satisfactorily his type of reaction. The determination of kinetic models for the thermal cracking process is of utmost importance for the development of experimental approaches based on numerical simulations, allowing then studies on the scaling-up process. The determination of the models, taking into account the mass balance of the process, as well as the reaction mechanisms involved, enables a better understanding of the process.