The switch from beryllium (Be) to tungsten (W) first wall (FW) armor in the new ITER baseline Pitts et al (2025 Nucl. Mater. Energy42 101854), Loarte (n.d. Plasma Phys. Control. Fusion) will impact the limiter start-up phase due to the much higher radiated power associated with W impurity. To assess this impact, a large database of SOLPS-ITER simulations of the hydrogen (H) plasmas in contact with the tungsten (W) limiter has been constituted, including transport of all individual W charge states up to a given cut-off. It has been shown that the plasma-limiter system exhibits strong self-regulating properties due to the strong dependence of W self-sputtering on the electron temperature at the last closed flux surface (LCFS). This self-regulation limits the power (PLCFS) crossing the LCFS into the scrape-off layer (SOL), resulting in high core radiated fractions (frad), but at the same time reducing the FW heat loads. Regression within the database also permits the derivation of a simple scaling relationship between the plasma temperature and density at the LCFS and the heating power which is used to construct a boundary condition for time-dependent scenario simulations. Simulations of W prompt redeposition using the ERO code for selected plasma backgrounds in the database found that 25%–45% of the eroded W particles reside in plasma for less than a gyration time. The reverse effect of the prompt redeposition on the background solution, assessed by re-running SOLPS-ITER with reduced W sputtering yield, proved have a greater effect on the solution than the impact of SOL plasma transport variation. Self-consistent simulations with an appropriate prompt redeposition model are therefore required to improve the prediction for ITER and to validate the code against W limiter start-up experiments on current devices. To assess the consequence of the inevitable background impurities present in real tokamak start-up plasmas, some preliminary simulations have also been performed with feedback-controlled nitrogen (N) seeding. It is found that even moderate N content leads to saturation of PLCFS, and hence frad, with respect to plasma density variation. The plasma density increase with seeding leads to redistribution of radiation between the W and N, preferentially cooling the edge in comparison with pure plasmas.