EVALUATION OF AN IMPROVED CODE FOR THE PERFORMANCE OF AFBCS

In a previously developed system model, bed and freeboard models were coupled for the continuous combustion of lignite particles of widely v aried size distributions, burning in their own ash in a fluidised comb ustor. This was modified to incorporate a) a procedure for faster comp utation of particle-size distributions (PSDs) without any sacrifice of accuracy; b) an energy balance on char particles, for determining tem perature variation with particle size; and c) assumption of plug flow for the interstitial gas. The improved computer code was evaluated by being applied to prediction of the behaviour of a pilot-scale fluidise d-bed combustor, followed by comparison of its predictions with measur ements and also with the results predicted by the previous code. The c omputer code replaces the conventional numerical integration of the an alytical solution of population balance with direct integration in ODE form, by using a powerful integrator LSODE; this results in a reducti on of CPU time by two orders of magnitude. For the prediction of physi cally expected variation of char hold-up with excess air, an energy ba lance on char particles must be incorporated into the system model. Co mparison of the predicted and measured temperature and concentration p rofiles shows that the present code produces better agreement than the previous one in bed-concentration profiles.