Channel Evolution of Sandy Reservoir Sediments Following Low-Head Dam Removal, Ottawa River, Northwestern Ohio, U.S.A.

Dozens of low-head dams are removed annually for reasons of obsolescence, financial liability, public safety, or as part of aquatic ecosystem restoration. Prior to removing a dam, hydrologic and sedimentologic studies are used to predict channel changes that would occur after the proposed dam removal. One commonly used predictive approach is a channel evolution model (CEM). However, most CEMs assume that the reservoir has trapped cohesive silts and muds. This study looks at the effects of low-head dam removal on a reservoir in filled with sand-rich sediment. The Secor Dam (2.5 m tall, 17 m wide) was constructed on the Ottawa River in northwestern Ohio (USA) during 1928 and was removed in 2007. High resolution channel cross-sections were measured at 17 locations prior to dam removal and re-measured every approximately 30 days for 6 months following the removal. Sediment sampling, sediment traps, substrate sampling, differential GPS tracking of channel bed forms and sediment coring were also used to characterize the channel sediment response to dam removal. Breaching of the dam produced a diffuse nickzone which was the width of the channel and about 10 m in length. One initial response was downstream migration of a sediment wave at rates up to 0.5 m/hr. The overall effect was erosion of the former reservoir to a distance of 150 m upstream of the former dam. Portions of the former reservoir were incised >1 m. Within the first 6 months after removal, approximately 800 m3 of sand had been mobilized from the former reservoir, transported downstream past the former dam, and had primarily in-filled pre-existing pools within a reach approximately 150 m downstream of the former dam. This behavior significantly differs from the predicted results of current CEMs which anticipate a first flush of suspended sediment and minor deposition of bed load materials in the channel downstream of the former dam.