A reduced-order model of char burning is developed for CFD modeling suitable for typical pulverized coal boiler. Comparing with the detailed one-dimensional model, the reduced-order model is much simpler and "smaller" but can still describe many aspects of porous char burning accurately by using the Thiele modulus method. The accuracy of variations of char conversion, particle size, apparent density and specific surface area predicted by the reduced-order model is firstly verified by comparison with the detailed one-dimensional model. Then combustion test data of a kind of anthracite conducted in a drop tube furnace are used to validate it further. The results indicate the reduced-order model can predict the char burning rate quite accurately with considering the ash inhibition effect and the predicted evolution of particle mass average diameter with mass loss agreeing well with experiment data. If comparing in detail by investigating particle size distribution carefully, the diameter change originated from chemical reaction can be described reasonably while finer particles generated due to fragmentation still could not be predicted by the current model. Finally, the reduced-order model is successfully implemented into CFD software, Fluent, and applied to a parallel CFD simulation of a 600-MWe supercritical W-shaped flame boiler. The reduced-order model can reasonably describe different particle size variation modes for different initial particle sizes in different temperature ranges. (C) 2015 The Combustion Institute. Published by Elsevier Inc. All rights reserved.