An Example of One River's Response to a Large Dam Removal

Damming a river can provide hydropower, water storage, and/or flood control, but at the same time may have adverse effects on fish and other aquatic species, flood land behind a dam, or degrade a river downstream of a dam. Many dams are undergoing a required relicensing process, during which time a dam's particular advantages and disadvantages are weighed. One way to mitigate the negative effects of dams is to remove them. However, dam removal itself can be problematic, especially when dealing with sediment stored behind the dam, which erodes and moves downstream.

Marmot Dam

The monitoring of the 2007 removal of Oregon's 15-m-tall Marmot Dam from the Sandy River (Figure 1), is a case study that examines the effects of removing a dam and releasing stored sediment. During the relicensing process, it was decided that the costs associated with maintaining and upgrading the Marmot Dam were higher than the benefits it generated and it was selected for removal. At the time of its removal, the 94-year-old dam was one of the tallest, and largest (in terms of stored sediment) dams to be removed in the United States. Documenting the response of the Sandy River provides valuable information for future large dam removals. This vignette provides a short description of the monitoring efforts on the Sandy River near the Marmot Dam. The data collection techniques are detailed in another vignette: (Sediment) Accounting 101: An Example.

Sediment

Rivers carry water and sediment [rock fragments such as sand and gravel], and dams affect the transport and storage of both. The Sandy River transports sediment ranging in sizes from fractions of a millimeter to half of a meter or larger. These sediment grains are carried suspended in the water column [suspended load] and along the bed of the river [bedload]. As the suspended load and bedload of a river encounter the slower, deeper water behind a dam, some or all of the particles stop moving and deposit on the river bed. Over the life of Marmot Dam, approximately 750,000 m³ of sediment had accumulated behind the dam, filling it nearly to the brim. At the time of removal, it was believed that this amount of sediment was roughly equivalent to 5-20 times the amount of sediment the river moved annually. The impounded sediment could be visualized as a 2.5 km long upstream-tapering wedge 15 m thick at the dam. Once the dam was removed, this reservoir sediment was once again subject to transport by the river, and had the potential to dramatically change the character of the Sandy River.

Upstream Erosion

The dam removal process involved constructing a small cofferdam [temporary dam] 70 meters upstream of the dam, removing the concrete dam with heavy equipment, leaving just the cofferdam atop the impounded sediment, and then breaching the cofferdam as the winter high flows arrived. Erosion started immediately upon its breaching on October 19, 2007. Within hours, the river had cut down several meters through the cofferdam and into the reservoir deposit. While the initial response was incision [the river cutting down through the sediment], the river rapidly migrated laterally [moved back and forth eroding banks] and cleared the sediment at the former dam site from bank to bank (Figure 2). A longitudinal profile of the river [elevation plotted versus downstream direction] shows the upstream progression of erosion (Figure 3). This erosion occurred by a combination of three processes: incision, lateral migration, and knickpoint migration [the moving upstream of a steep break, or step, in the bed slope]. Within one month, 25% of the reservoir sediment had been eroded, and by the end of the first year 50% had eroded. Two years after the dam was removed, the river had a similar width and similar banklines to its pre-dam state, but the bed elevation was still slightly higher than it was before the river had been dammed.

Downstream Deposition

Much of the eroded sediment did not travel far. As a result of the sediment released from the reservoir, the bed of the river immediately downstream of the former dam aggraded [increased in elevation] by as much as 4 m. The river initially changed from a single-channel river to a multiple-channel one due to the sediment introduced from upstream (Figure 4). By the end of the first wet season, the sediment immediately below the dam stabilized into one new large bar and the river went back into a single-channel configuration. The small sediment sizes (<2 mm) moved down the river as suspended load. Much of the larger sediment (2 mm-20 cm) was deposited in the 2 km immediately downstream of the dam. Very near the dam, the sediment deposited uniformly across the river; cross sections show this section-wide aggradation (Figure 5). However, farther downstream the sediment deposited less uniformly--primarily on top of existing gravel bars. Most of the monitoring effort focused on the reach [portion of the river] near the dam, where the large changes occurred; annual surveys 14 and 20 km below the dam failed to show change in the two years after the dam was removed.

Overall Observations

The Marmot Dam removal provided an opportunity to monitor the response of a portion of the Sandy River to a sudden increase in the sediment supplied to it. The reservoir bed showed three erosion processes--incision, widening, and knickpoint migration. The downstream deposition differed based on grain size and proximity to the dam. Much of the sediment quickly stabilized into the bed near the dam. Over the remainder of the river there was no measurable change in the river bed. This response is similar to pre-removal predictions using a numerical sediment transport model (Cui and Wilcox, 2008). The multi-year monitoring project has documented how the Sandy River has processed the large amount of sediment that was made available to it by the removal of the Marmot Dam and should help inform future dam removal decisions.

Cheng, F., and Granata, T., 2007, Sediment transport and channel adjustments associated with dam removal: field observations: Water Resources Research, v. 43, W03444, doi:10.1029/2005WR004271, 14 p.

Cui, Y., and Wilcox, A., 2008, Development and application of numerical models of sediment transport associated with dam removal, in Garcia, M.H., ed., Sedimentation Engineering--Processes, Measurements, Modeling, and Practice.American Society of Civil Engineers Manuals and Reports on Engineering Practice no. 110, p. 995-1020.

Doyle, M.W., Stanley, E.H., and Harbor, J.M., 2003a, Channel adjustments following two dam removals in Wisconsin: Water Resources Research, v. 39, no. 1, 1011, doi:10.1029/2002WR001714, 15 p.

Doyle, M.W., Stanley, E.H., Harbor, J.M., and Grant, G.E., 2003d, Dam removal in the United States: Emerging needs for science and policy: EOS, Transactions American Geophysical Union, v. 84, p. 29, 32-33.

Heinz Center, 2002, Dam removal--science and decision making. The H. John Heinz Center for Science, Economics, and the Environment, 221 p.

Major, J.J., O'Connor, J.E., Grant. G.E., Spicer, K.R., Bragg, H.M., Rhode, A., Tanner, D.Q., Anderson, C.W., Wallick, J.R., 2008, Initial fluvial response to the removal of Oregon's Marmot Dam: EOS, Transactions American Geophysical Union, v. 89, no. 27, p. 241-242.

National Inventory of Dams, 2007, US Army Corps of Engineers. https://nid.usace.army.mil/. accessed 1 Dec 2009.

O'Connor, J., Major, J., and Grant, G., 2008, The dams come down--unchaining U.S. rivers: Geotimes, v. 53, no. 3, p. 22-28.

Pizzuto, J., 2002, Effects of dam removal on river form and process: BioScience, v. 52, p. 683-691.

Stillwater Sciences (2000), Numerical Modeling of Sediment Transport in the Sandy River, OR Following Removal of Marmot Dam, Technical Report, prepared for Portland General Electric, Berkely, CA. available online at: http://www.stillwatersci.com/publications.php

• Johns Hopkins University/National Center for Earth-surface Dynamics: http://www.jhu.edu/marmot/
• US Geological Survey: http://or.water.usgs.gov/projs_dir/marmot/index.html