Water plumes erupting from the 'tiger stripe' features on the south pole of Enceladus are thought to connect to a global subsurface ocean. Proposed origins for the initial stress necessary to form the 'tiger stripes' include a giant impact, which would require true polar wander, or tensile stresses, which would require a partial freezing of the subsurface ocean. A further issue with these hypotheses is that the 'tiger stripes' may be short-lived. We show here that impact resurfacing can seal off plumes and mass loss can lead to their compression and closure over ∼ 1 Myr. Since plumes are observed at present, a mechanism by which new plumes can be generated every ∼ 1 Myr and by which such plumes are most likely to form at the south pole is needed. We propose and investigate the possibility that impacts constitute a adequate repeating source for the continual instigation of fractures and plumes. We find that the rate of impacts on Enceladus suggests the formation of ∼ 10 3 independent plume systems per Gyr, the vast majority on the south pole, and is consistent with the Cassini-derived age of the south pole for a lunar-like bombardment history, our estimates of fracture lifetimes, and with the needed parameters for parallel fracture propagation. The model favors a bombardment history similar to that of Triton over one more similar to that of the Galilean satellites, and favors a cumulative power-law index of 4.2 for impactors with radius, 1 km < R < 10 km.