We introduce a maintenance model for a single component that is part of a complex engineering system and has a monotic, stochastic degradation process. For this component, a warning limit is given and we use a condition-based maintenance (CBM) policy. The other components in the system are assumed to follow failure-based and periodic maintenance policies and their maintenance actions lead to unscheduled and scheduled downs of the system. These downs constitute opportunities for the single CBM component. If the degradation is at or above a control limit at an unscheduled or scheduled down, the component is replaced preventively and one saves on downtime and setup costs. We derive an efficient and accurate approximate evaluation procedure. Further, we show that this approximate evaluation leads to a close-to-optimal control limit policy when used within the optimization procedure and we show the potential savings when both unscheduled and scheduled downs are used as opportunities for the CBM component. Finally, we demonstrate that our model can be well used as a building block to solve multi-component CBM problems with many components.