Enhancement of solar-to-electric efficiency in future solar power tower (SPT) plants will require an improvement of the power cycle performance. An attractive option is the use of a supercritical CO2 Brayton cycle, which needs a heat transfer fluid (HTF) temperature of around 700 °C to be competitive. However, the current HTF employed in SPT plants, i.e. nitrate salt, cannot reach this temperature target. For this reason, new HTFs with a large temperature operation range have been investigated recently, but usually their freezing temperature is comparatively high, which complicates the preheating of the tubes during the receiver start-up. In the preheating operation, the heliostat field concentrates the solar irradiation onto the receiver tubes to elevate their temperature above a minimum value, while they are still empty, to avoid the risk of freezing of the incoming flow of HTF. This is a very critical operation, as too much incident heat flux in the preheating leads to extremely high thermal-stresses, causing the creep-fatigue damage of the receiver tubes. The present work numerically characterizes the temporal evolution of the temperature and thermal-stresses during the preheating of a typical configuration of a receiver tube made of Haynes 230 alloy. The aim is to check whether it is possible to preheat the receiver tubes above the freezing temperature of a chloride salt (i.e. NaCl–KCl–MgCl2) as an alternative high-temperature HTF to the nitrate salt. The results indicate that, although the original Vant-Hull algorithm of tube preheating is suitable for the nitrate salt, it is not adequate for the chloride salt. Therefore, a series of modifications of the Vant-Hull algorithm are proposed to meet the requirements of the high-temperature HTF. In the modified Vant-Hull algorithm, the resulting peak values of the temperature gradient and the von Mises stress are reached within the first 1.2 min of preheating. Besides, in the cases analyzed, the estimated creep-fatigue damage of the tube produced with the modified algorithm is much lower than an allowable limit set to protect the structural integrity of the receiver. However, the demanding operating conditions of the chloride salt receiver lead to a lower lifespan than that of the conventional nitrate salt receiver for the same heliostat field and receiver layout. Furthermore, the modified Vant-Hull algorithm requires longer preheating times, which delays the start-up of the SPT plant and reduces its available hours of operation per day. Finally, this work also demonstrates that the modified Vant-Hull algorithm is suitable to preheat receiver tubes that operate with other high-temperature molten salts and liquid metals.