A three-step methodology was developed to provide reliable prediction of a coal's behaviour in a utility boiler: (1) Extracting the combustion kinetic model parameters by combining experimental data from a pilot-scale test facility, Computational Fluid Dynamic (CFD) codes and an artificial neural network. While the combustion kinetic parameters used in the model code will not correspond to the combustion rate of a single particle of coal, these parameters do describe the combustion behaviour of a "macroscopic" sample of tested coal.(2) Validation of the combustion kinetic model parameters by comparing diverse experimental data with simulation results calculated with the same set of model parameters. (3) The model parameters are then used for simulations of full-scale boilers using the same CFD code. For operational engineering information needed by the utility operator, the authors apply the predicted results to EXPERT SYSTEM, a boiler supervision system developed by Israel Electric Corporation (IEC). Four different bituminous and sub-bituminous coals with known behaviour in IEC 550MW opposite-wall and 575MW tangential-fired boilers were used to show the adequacy of the methodology. The predictions are done with the CFD code, GLACIER, propriety of Reaction Engineering International (REI). Preconfigured GLACIER models of the test and full-scale furnaces were purchased from REI and validated by our group. This book includes a detailed description of the methodology, test furnace facility and an example of the experimental and predictive combustion results from the four coals used to test the methodology. In addition, two previously unknown coals are examined prior to their firing in the utility boilers and prediction of their behaviour and operational parameters in the two boilers carried out.