Hybrid systems consisting of a single or multiple primary sources along with a storage component are used in renewable energy production due to the wide disparity between the intermittent power generated and the power demand. Finding the optimal size of a hybrid system with no loss of power supply (LPS) is of utmost concern when considering the levelized cost of energy (LCE) of the project over its lifecycle. In this study, an optimization routine employing a search algorithm is developed to find the minimum LCE that meets zero LPS in the context of data center application. To this end, a system model is developed by integrating basic models of the subsystems. The system model is then used to investigate two different loading cases, 1) a case where the demand cannot be controlled as in the case of power demand in a residential network, and 2) a case where the demand can be controlled up to certain limits, as in the case of power demand in a data center or a data center network. Various types of controllable power demand scenarios are studied. The results demonstrate a significant reduction in the life cycle costs of the system when controllable power demand is considered, especially in regions where wind and solar resources are relatively low. (C) 2012 Elsevier Ltd. All rights reserved.