Batch experiments were conducted to investigate the emissions of carbon oxides and nitrogen oxides from the combustion of char prepared from a commercial coal in a bubbling fluidized bed. Combustion gases escaping from the surface of the bed were sampled and analyzed for N2O, NO, NO2, CO and CO2 as a function of time by means of a Fourier Transform Infrared Spectrometer. The experimental variables include char size and loading, inlet oxygen concentration, inlet CO concentration and bed temperature. A single particle reaction-diffusion model was developed first. The detailed chemistry of NO and N2O formation destruction is complex. Homogeneous and heterogeneous reactions considered most pertinent were included in an extension of the single particle char combustion model reported earlier. This single porous char particle model was integrated into a three-phase hydrodynamic description of the fluidized bed reactor. This system model for fluidized bed combustion was developed in response to the observation that most previous system models, including those that incorporate details of NOx emissions, are based on the two-phase theory of fluidization. However, two-phase models are unable to predict the gas back-mixing and the recycle peak in solids-mixing. The non-isothermality of the bed resulting from the gas-phase reactions was taken into account through inclusion of an energy balance for the bubble phase. The effect of the variation in superficial gas velocity on bubble properties and cross-flow was included through an overall mass balance. Calculations from the system model, including details of homogeneous NOx reactions far from the char particles, compared well with data on the emissions of CO, CO2, NO and N2O for various experimental conditions. The validated model was then used to investigate the influence of operating conditions on the conversions of char-nitrogen to NO and N2O from the simultaneous combustion of char and propane in a fluidized bed with an air/propane mixture as the fluidizing medium. The most significant factors were found to be bed particle size and char diameter. (C) 2001 Elsevier Science Ltd. All rights reserved.