The 1994 seepage piping failure of IVEX dam, resulting in catastrophic dewatering of the reservoir, was an analog for the one-step process commonly used for dam removals. This study evaluates the impact of the sediment release in the context of channel evolution models (CEMs), and compares short-term (2-month) versus long-term (12-year) released sediment volumes using a downstream reservoir that acted as a sediment trap. The CEM of Doyle et al. (2003) calls upon six stages (stage A-F) in the restoration of quasi-equilibrium conditions following a dam removal. This study confirmed the model's validity but proposes several significant changes: (1) a new stage A2 represents pre-dam failure sediment erosion including channeling and longitudinal scours in the reservoir sediments; (2) an extensively modified stage B represents development of an early-breach drainage network that cross-cut earlier features; and (3) an extensively modified stage E represents lateral channel migration and incision, resulting in channel backfilling (terraces and point bars). The 12-year study shows that respective stages lasted from minutes (stage B), to days (stage C), to 2–3 months (stages D and E), to 3–5 years (stage F). Failure of IVEX dam had significant impacts on the downstream reservoir. For the 2-month interval following failure, sedimentation rates increased about three orders of magnitude, and the volume of trapped sediment approximately doubled. However, over the subsequent decade this initial sediment flush was only about one-third of the decadal sediment loading, illustrating the importance of long-term downstream migration of the coarse-grained sediment fraction stored in the reservoir.
Journal of Great Lakes Research
Vol. 33 • No. sp2
Vol. 33 • No. sp2