The paper presents new Monte-Carlo simulations of the transport of 13CH4 methane injected through a hole in a testlimiter positioned at the last closed flux surface of TEXTOR. The results show that the spatial distribution of 13C re-deposited locally on the testlimiter surface can be modelled if the parameter S for the sticking of returning hydrocarbons 13CHy is set to zero or at least almost zero. This is interpreted as a negligible effective sticking probability of the returning hydrocarbon radicals. The re-deposited C-species are directly re-eroded and disappear whereas a direct reflection seems to be unrealistic. A re-erosion might be caused by the hydrogen carried with the CHy radicals ("self re-erosion"). The calculated deposition efficiency of 13C at the testlimiter surface, however, remains much too high compared with the observed one (less than 0.5%). A low deposition efficiency can only be modelled if in addition an enhanced yield for chemical erosion caused by the background hydrogen (Ye ~ 8% compared to Y ~ 2% for graphite under the given conditions) for the fresh redeposits is assumed. Similar assumptions reproduce the high amount of carbon deposition observed on the inner louvers in the MkIIa divertor configuration of JET and on the plasma-shadowed areas of the MkIIGB divertor. A possible explanation could be a synergistic effect of the hydrogen ions from the plasma background leading to an enhanced erosion by influencing the properties of the redeposits. In contrast to carbon, modelling of the beryllium transport in the divertor of JET MkIIa shows no significant deposition on the louvers. This is in good agreement with experimental findings. Beryllium deposited at areas with low electron temperatures is not eroded and transported anymore because the hydrogen ion impact energy is below the threshold for physical sputtering and beryllium does not suffer from chemical erosion.