This paper presents and investigates a novel, hybrid model predictive approach to control the auxiliary heating for a combistorage. Faulty design, as well as suitable design schemes cause situations in which solar energy supply has to "compete against" the auxiliary energy supply. This research demonstrates a feasible method to remedy such situations with the utilization of weather forecast data. The developed approach is modular and expandable to be used with additional heat sources. A suitable disturbance-prediction, which approximates the expected solar energy flux into the storage, in connection with a linear model predictive control (MPC), can prevent the auxiliary system to switch on at an early stage and thus reduce the auxiliary energy demand and keep storage capacity for solar thermal energy supply. Results obtained through simulations for selected months show a reduction of auxiliary energy demand up to 40%, when facilitating this approach for a solar thermal combisystem for a single family house. Monthly solar fractions (F-s) increase by approximately 4% points or 5% with respect to the base case. (C) 2013 Elsevier Ltd. All rights reserved.