Many applications in wireless sensor networks require an absolute predictable network behaviour under all possible traffic and routing conditions. This requires network dimensioning with respect to crucial parameters such as energy usage, message delay and buffer requirements. Network calculus provides a method set that can be used to prove that all possible traffic conditions under a given routing-implied topology can be supported. As it is often impossible to fix the routing topology before or even during the operation of a sensor network, an uncertainty arises with which network calculus cannot deal directly. In this paper, we make two contributions: 1) We demonstrate how to apply network calculus in a fixed sensor network topology. 2) We show how to incorporate topology uncertainty into these calculations.