The increasing costs of energy encourage the development of cost-efficient building cooling systems. Based on a thermal model of a commercial building and its cooling system, the cost-optimal design of a vapor compression system and an incorporated ice storage is determined. Special emphasis is placed on the refrigeration machine and its interaction with the thermal capacity of the building. The implementation as a mixed integer linear programming problem allows the evaluation of the optimal system operation for an entire year in approximately 50s on a desktop computer. Based on these results, the design of the optimal systems is determined for various electricity tariff schemes. Therefore, a model for variable electricity rates is introduced including costs for control reserve power. Compared to the cooling system without storage, the system including an optimally designed ice storage achieves lifetime cost reductions of approximately 8% by reducing the operational costs and the investment costs for the downsized refrigeration machine. (C) 2012 Elsevier B.V. All rights reserved.