(690e) Parametric Analysis of Brown Coal Pyrolysis in a Drop Tube Furnace Using CFD-DPM
AIChE Annual Meeting
2015
2015 AIChE Annual Meeting Proceedings
Particle Technology Forum
Fluidization and Fluid-Particle Systems for Energy and Environmental Applications I
Thursday, November 12, 2015 - 1:50pm to 2:10pm
A Three-dimensional unsteady state mathematical model is developed to numerically investigate the effect of different parameters namely, the operating temperature, solid mass flow rates, coal particle diameter on the pyrolysis of air-dried Victorian brown coal in a lab scale drop tube furnace (DTF) under pure nitrogen atmosphere. The DTF is an alumina cylindrical reactor with 3.8 m long high temperature electric flow furnace known as 'Helena' operated with a maximum temperature of 1600 0C at ambient pressure. Coal particles are fed along with the primary gas at the centre of the reactor. Secondary gas is also introduced through an eccentric port at the top of the reactor. The solid product from pyrolysis is collected at the base of the reactor for offline analysis.
Eulerian Lagrangian modeling approach is followed to simulate the pyrolysis process. The model includes the physical and chemical phenomena such as the hydrodynamics, heat transfer and chemical reaction to determine the emission of volatile matter from parent coal. Six species (N2, CO, CO2, H2, CH4 and VM) in the gaseous phase are considered in the model. It is assumed that the mechanism of pyrolysis involves two steps. First, the volatile matter (VM) as a whole emits from the coal particle during pyrolysis and subsequently gets decomposed in to various gaseous products (CO, CO2, H2 & CH4) through volumetric reactions in the fluid phase. It is further assumed that the devolatilization process as well as the gas phase decomposition reaction follows a first order global reaction model with Arrhenius kinetics. The rate constant of the decomposition is selected such that the process gets completed within negligibly small fraction of the simulation time step. Fluid phase turbulence is captured using standard k-ε model proposed by Launder and Spalding [1], drag between gas and coal particles are determined by the correlation proposed by Wen & Yu [2] and radiation heat transfer between the wall and particles is simulated by discrete radiation model.
Mass flow rates of the primary and secondary streams as well as coal particles are specified at the inlet sections of the reactor. At the outlet of the reactor, the pressure is specified to be atmospheric. The primary gas temperature is assumed to be ambient (300 K) and the secondary gas temperature is 733.15 K according to experimental condition. Usual no slip condition is applied at the walls for gas phase. The walls are considered to be elastic for solid particle interactions. The side wall is maintained at constant temperature. The upper wall of the furnace is considered to be adiabatic. It is assumed that the reactor contains pure nitrogen the beginning of the process.
Simulations are carried out using rate constants obtained from present thermogravimetric experiment with activation energy 274 kJ/mol and pre exponential factor or Arrhenius constant 2.7E-9 s-1, for different operating temperatures, coal mass flow rate and particle diameters using finite volume method as implemented in the commercially available software ANSYS Fluent 14.0 . First order upwind discretize method is used to convert the non-linear differential equation to linear algebraic equation. The pressure linked equations are solved using SIMPLE algorithm. An absolute convergence criterion of 10-8(in appropriate units) is set for the simulations. Grid convergence test are carried out to optimize the computation time.
Results show that the rate of pyrolysis increases with increase in temperature. It is also observed that the particle residence time increases marginally with the increase in temperature. The steady state temperature distribution in the reactor seems to be independent of the feed rate as well as the size of the coal particles used in the simulations; however, the initial rate of devolatilization is faster as the particle size decreases.