Numerical modelling of combustion processes at elevated pressures

As more and more CO2 quantities are discharged into the atmosphere from flue gases due to fossil fuel combustion, increasing concerns over greenhouse gas emissions have caused extensive research to be directed to the field of new power generation cycles that enable CO2 capture and storage. Raising the pressure of the coal conversion enables economic separation of CO2. Two technologies working at elevated pressures are, for example, the Integrated Gasification Combined Cycle (IGCC) technology and the pressurised oxy-fuel combustion power cycles. Both technologies rely on conversion of coal at elevated pressures, and hence enabling the CO2 separation process. For the numerical simulation of such processes, it has been proven that global models of atmospheric char conversion are neither directly applicable nor extrapolatable for elevated-pressure atmospheres. Therefore, developing more reliable char gasification/combustion models is a key point in being able to successfully predict the char gasification/combustion processes taking place at elevated pressures for these new clean-coal technologies. Since oxygen is the most efficient oxidising agent, the principal aim of this work is to develop a mechanistic model for Char Burn-Out (CBO) that is capable of providing correct predictions of char oxidation rates for wide ranges of temperature and pressure. In order to validate the predictions of the model, the finite volume (FV) CFD combustion simulation code AIOLOS has been used. During the course of development of the coal conversion model, a SIMPLEC algorithm (Semi Implicit Method for Pressure Linked Equations Consistent) for non-staggered (collocated) grids has been developed, since its robustness has been proven to be higher than their SIMPLE counterpart. An elevated-pressure pyrolysis model has been introduced, where a single first order reaction kinetic rate has been adopted to predict devolatilisation of raw coal. The total volatile yield prediction has been achieved by creating a data base of 14 coals, on the van Krevelen (coalification) diagram, which range from low volatility to lignite coals, and whose total volatile yields are fixed by experiment. An interpolation technique has been adopted afterwards to predict the total volatile yield of coals which are not directly obtainable from the data base. Tar yields are predicted by means of a model that correlates the tar yield to the elemental composition of the parent coal and to the system pressure applied. The mechanistic char burn-out model relies on a three-step char oxidation process in contrast to the well-known two-step chemisorption-desorption mechanism (Langmuir-Hinshelwood kinetics). The third reaction step represents a reaction between oxygen molecules and the already formed carbon-oxide complexes. The specific reaction rate of char oxidation has been modelled by means of the Extended Resistance Equation (ERE) which accounts for Boundary Layer Diffusion (BLD) effects as well as pore diffusion effects by using a reaction penetration factor that accounts for the contribution of the interior surface of the coal particle to the oxidation process. The predictions of the two-step ERE and the three-step one have been compared against the results of experiments conducted on a pressurised Entrained Flow Reactor (p-EFR). The comparison has reflected the successfulness of the three-step char oxidation mechanism of predicting a combustion rate enhancement as partial pressures of oxygen are increased in contrast to the two-step model.