Many studies have been on the effective heat conductivity (λeff) and the transfer coefficient at the wall (αw) inside packed beds. It has been mentioned that the values of λef and αw are changed when a chemical reaction occurs in the packed bed. We give an explanation for such a phenomenon. The properties λeff and αw are lumped parameters which usually are determined by both the measured temperature profiles and the model used to calculate the temperature profiles from λeff and αw. If either the experimental data are wrong or the model is erroneous the error will manifest itself in the values of λeff and αw. At least a part of the change in the values of λeff and αw due to a chemical reaction is caused by the fact that a homogeneous model with catalyst and gas having the same temperature is chosen, whereas a heterogeneous model with catalyst and gas having different temperatures should be used. If no reaction occurs the catalyst and gas will have the same temperature and the homogeneous model yields a good description. Hence, when fitting temperature profiles with this model the correct values of λeff and αw are found. If reaction does occur the catalyst and the gas will have different temperatures because the heat of reaction must be transferred from the catalyst to the gas. If, despite this fact, a homogeneous model is used to calculate the temperature profiles, an error is introduced which is reflected in the values of λeff and αw. As a consequence we create an apparent dependence of λeff and αw on the reaction rate. We derive criteria to determine which model must be used. We discuss results presented in the literature on the dependence of λeff and αw on the chemical reaction. The explanation is both qualitative and quantitative.