A dynamic model for a fixed-bed reactor for methanol synthesis is presented. The model is compared with its steady state version. The analysis points out that the numerical stability of the dynamic model is improved by opportunely increasing the level of detail. It is appropriate to introduce the diffusion terms, to work with mass fractions, to select good discretization methods for each term of the model equations. Since these aspects are usually neglected in steady state analysis, this paper investigates step-by-step their implementation, emphasizing their importance (I) in the transformation of an original hyperbolic PDE system into a parabolic PDE system; (II) in removing non-physical oscillations generated by first-order systems that may lead to relevant model prediction errors; and (III) in the approximation of the convection terms using the forward formulation, which is more stable and provides more realistic solutions. (c) 2012 Elsevier Ltd. All rights reserved.