A key feature of the German "Energiewende" is a shift towards variable renewable power sources like wind and photovoltaics (PV). According to the medium-term scenario of the network development plan drafted by German TSOs, onshore and offshore wind could account for around 45% of gross power demand by 2032, whereas PV could contribute around 10% (NEP 2012, scenario 2032B). Afterwards, the shares of wind and solar are projected to grow further until 2050 (DLR et al. 2012). The hourly feed in of both wind and PV is only weakly correlated with hourly load profiles. Growing shares of these technologies thus have a strong influence on residual load, for example resulting in temporary situations of power shortage or renewable surplus generation (Consentech and r2b 2010, Denholm and Hand 2011). Integrating growing amounts of wind and PV into the power system increasingly requires the application of dedicated integration measures, for example storage, demand-side measures, network expansion, conventional back-up and renewable curtailment (Dena 2010, Steffen and Weber 2013,VDE 2012a). In this paper, we focus on renewable surplus generation, storage and curtailment and aim to answer two research questions. First, we are interested in the future development of German residual load under a range of varying assumptions. In particular, we analyze the nature of renewable surplus generation (power, energy, and duration). Second, we determine how much storage of different technologies would be required for taking up temporary renewable surpluses. We specifically explore the interrelation of storage and renewable curtailment: how are storage requirements reduced if increasing levels of renewable curtailment are tolerated? Noticeably, the analysis includes a large number of sensitivities. We consider different renewable expansion scenarios, different developments of load and must-run restrictions, various meteorological wind and PV years, as well as different levels of biomass flexibility.