Sediment Strategies: Choosing a Sediment Management Option for Dam Removal

Four options were considered for managing sediment during the removal of a large dam in the western U.S. The option chosen – called the retained sediment method – involves leaving sediment in place and rerouting the river around the dam site.

By Peter L. Stroud

Over the past decade, 20 to 50 dams have been removed annually in the U.S., according to American Rivers, a group that works to protect and restore rivers and streams in the country. The reasons behind these removals vary. This decision may be made during the Federal Energy Regulatory Commission (FERC) relicensing process of a hydroelectric facility or to address a need for fish habitat restoration. Dam removal may be needed to deal with potential hazards during floods or seismic events or may be required due to structural deficiencies.

Sediment management plays a critical role in the choice of a dam removal method, as sediment trapped behind a dam drastically affects the cost and environmental impact of the removal. Released sediment – often millions of cubic yards – can bury downstream habitat, threaten infrastructure, pose water quality concerns, change channel dynamics and increase flood hazards.

Every dam removal effort is unique. Yet the process to reach a decision about removal or reinforcement follows a common evaluation process – particularly when it comes to sediment management – and is based on lessons learned in the past decade.

The work currently being done to further the planned removal of a large dam in the western U.S. provides a comprehensive look inside the complex decision-making process that goes into sediment management when a dam is being removed or considered for removal.

A notch/release approach to sediment management during removal of Glines Canyon Dam allowed 13 million cubic yards of sediment to move downstream in controlled increments, creating less turbulence and disruption to the environment and downstream infrastructure.
A notch/release approach to sediment management during removal of Glines Canyon Dam allowed 13 million cubic yards of sediment to move downstream in controlled increments, creating less turbulence and disruption to the environment and downstream infrastructure.

A dam in distress

In the early 1990s, it was determined that the dam being discussed in this article did not meet modern seismic stability or flood safety standards. In 2007, the owner of the dam conducted a comprehensive study of ways to ameliorate the safety issues, including strengthening the dam and removing the dam. In the end, the owner decided the most feasible alternative for dealing with this situation was to reroute a portion of the river, to allow the sediment to remain in place, and remove the dam. This would resolve the dam safety issues while also addressing the other environmental issues related to the dam’s impact on the river. Positive effects of removal will include restoration of a lagoon and floodplain, increase of instream flows and improve spawning habitat for trout.

But what about the sediment? The reservoir behind the dam is 90% filled, holding back about 2.5 million cubic yards of sediment.

Fundamentally, there are four dam removal and sediment control approaches: notch/release, rapid release, dig and dewater, and retained sediment. These approaches are driven by several factors, including the amount of time available to complete the work, the schedule of the project subtasks, environmental and infrastructure impacts to the reservoir and downstream areas, and cost. The removal work also needs to have logical stopping points during non-work periods if it extends beyond one in-water working period.

Sediment releases can fill channels downstream, changing flood patterns and burying infrastructure. Erosion of the dam upstream can create unstable slopes and also put infrastructure, such as bridges and pipelines, at risk. The duration of the removal project, the care taken to protect the environment and the approach to managing the sediment significantly affect costs.

The team involved in the work to remove the dam evaluated these four dam removal and sediment management methods with a goal of determining a best-fit strategy for removal.

The notch/release approach

One common method used to remove dams is called notch/release, in which the water level in the reservoir is gradually lowered to the bottom of the spillway gates. Layers of the dam structure are then removed, typically on alternate sides of the dam, to create temporary spillways or notches. The size of the notches depends on the flow required to lower the reservoir level. This approach releases the sediment in a slow, steady flow, which allows the downstream areas to adjust to the new materials, creating less turbulence.

The owner of the Elwha and Glines Canyon dams near Port Angeles, Wash., opted to use the notch/release approach during removal of these two dams. The dams were built to produce electricity on the Olympic Peninsula. Elwha Dam, constructed between 1910 and 1913, is a 105-foot-high concrete gravity dam. Glines Canyon Dam, built in 1927, is a 210-foot-high concrete arch dam.

The dams were operated in run-of-river mode and generated about 40% of the electricity needs for the Diashowa America paper mill in Port Angeles. However, the dams had no facilities for upstream passage of anadromous fish. Removal would provide an opportunity to restore an entire watershed – which is in Olympic National Park – to near-natural conditions.

The U.S. Department of the Interior purchased the dams in 2000 with the goal of removing them. Key in the removal plan was to determine how to manage about 18 million cubic yards of sediment trapped within the two reservoirs, 13 million cubic yards behind Glines Canyon Dam and 5 million cubic yards behind Elwha Dam.

Elwha Dam forms Lake Aldwell 8 miles upstream from the mouth of the Elwha River. Glines Canyon Dam forms Lake Mills 13 miles upstream from the river’s mouth. Because the majority of the sediment (about 13 million cubic yards) is stored in Lake Mills, experts developed a sediment erosion model to predict the rates and final quantities of sediment deposition. About half of the sediment in Lake Mills is coarse-grained (sand, gravel and cobble-sized), and half is fine-grained (clay and silt-sized). Some of the fine-grained sediments are transported through the reservoir, while the remainder of the sediment is deposited along the reservoir bottom. All of the coarse-grained sediments are trapped within the reservoir as delta deposits located at the upstream end. The width of these reservoir deltas is large – about 10 times greater than the width of the alluvial river channel.

The selected plan for managing the 18 million cubic yards of reservoir sediment was to remove both dams concurrently in controlled increments using the notching method.

Dam removal began on the Elwha River in September 2011. First, the water level of the reservoir was lowered. Then, a part of the dam was removed using hydraulic hammers. Removal proceeded quickly, and by late spring 2012, Elwha Dam was gone.

The work on lowering Glines Canyon Dam commenced in September 2011 and is expected to be complete by summer 2013.

Rapid release approach

The least expensive sediment management approach is rapid release. This approach may involve constructing a tunnel through the base of the dam, which is then opened, creating a catastrophic release of sediment and water. The procedure causes rapid erosion, deposition and scour, as well as extreme turbidity in the river. The rapid release approach has greater impacts to downstream aquatic life but shortens the duration of impact compared to the notch/release method.

It can also have greater impact to infrastructure. For instance, the rapid release can cause significant erosion and scour upstream, exposing foundation footings and buried pipelines, as well as flooding and deposition downstream, putting flood-control structures such as levees at risk, as well as facilities protected by the levees. However, if the downstream area does not have exposed infrastructure and the sediment can discharge into a larger river system with limited aquatic impact, this approach may be favored.

In 2011, a dam owner in the Northwest U.S. used the rapid release approach to remove a dam and create immediate fish passage. As compared to Elwha Dam, this dam is located several miles closer to the confluence of a major river system that has significantly higher sediment loads and lacks downstream infrastructure and population centers. Thus, the environment could tolerate the short-term turbidity spikes with limited impacts to aquatic life and the owner opted to breach the dam with a rapid release of sediment. Crews cleared sediment and debris away from the area immediately upstream of the 13-foot-by-18-foot drain tunnel and then blasted open the upstream end of the tunnel.

The reservoir drained within about 90 minutes and flushed an estimated 2.4 million cubic yards of sediment into the river. The release of sediments occurred as expected, without significant problems.

Dig and dewater approach

The most expensive sediment management approach – in terms of cost per yard of sediment – is to drain the reservoir, dig out the sediment and haul it off-site. Typically, the sediment will need to be dewatered before off-site disposal, which means the materials must be handled more than once. Although the engineering may be less complicated as compared to in-place stabilization of the sediment, there are significant costs associated with digging, dewatering and hauling.

The owners of Mill Pond Dam in northeastern Washington plan to use the dig, dewater and haul method to manage sediment behind the dam. In this case, the dam and its reservoir are not large. The dam is 134 feet long and about 55 feet high and is part of the 4-MW Sullivan Creek Hydroelectric Project, located within the Colville National Forest in northern Pend Oreille County. The dam is on Sullivan Creek 3.5 miles upstream from the Pend Oreille River, which is a tributary of the Columbia River. Boundary Dam is located 10 miles downstream from the mouth of Sullivan Creek on the Pend Oreille River. Boundary Dam is one of three dams between Mill Pond Dam and the Columbia River that could be affected by released sediment.

The dam removal is planned to take place in one construction season. The plan is to remove excess reservoir sediment, dewater it and dispose of it in upland areas. Also, coarse gravel and cobbles may be mined from the reservoir sediment and used to restore the stream channel and fish habitat. Although the project area is relatively small, the handling of the reservoir sediments will incur higher relative costs compared to allowing the sediment to be released downstream.

Retained sediment approach

The fourth management method is to retain the sediment in place. Commonly, the dams that are being removed are older and their reservoirs have been filled in. If the environmental consequences are of great concern and sediment cannot be released downstream, one option is to stabilize the reservoir sediment in place and channel the river around it. This approach may require extensive engineering. The project may also need to be done in phases over a two- to three-year period, and each phase will need to end with the project being in a stable state to withstand the non-work winter season.

A balanced sediment solution

Since the mid-1990s, multiple engineering and environmental studies have been prepared by the owner of the aforementioned dam in the western U.S. to determine what should be done to improve the situation for safety and environmental protection.

Three of the four options mentioned above would not work for this dam. Sediment release either through notching or rapid release would affect downstream habitats, estuaries, water supply and riverside properties. Currently there is a significant flooding issue along the lower portion of the river, and allowing the sediment to erode downstream would likely worsen downstream flooding. In addition, because the dam is in a remote location, hauling sediment would be difficult (and costly).

Thus, the project sponsor and dam owner opted for removal with in-place sediment stabilization and rerouting the river. The plan is to reroute a half-mile portion of the river into a creek and use the abandoned reach as sediment storage area, then remove the dam. The creation of a new stream would include step pools that are fish-friendly.

The estimated cost of this alternative is about $85 million. The option to reroute the river and remove the dam received a great deal of support from resource protection agencies, as well as environmental protection groups, which supported removing the dam because it serves as a barrier to certain species of trout. Removing the dam allows steelhead unimpaired access to more than 25 miles of natural spawning habitat.

Once the channel excavation and stream restoration is complete, the dam will be demolished. The concrete rubble will be used to help stabilize the sediment stockpile and diversion dam. No construction wastes will need to be trucked off-site.

The dam owner has solicited design-build proposals from construction/engineering teams to: bypass the river into the drainage of the adjacent creek, stabilize the sediment in place, remove the dam and construct a new fish-friendly channel from the base of the dam to the point of the upstream diversion. The work will be awarded in the spring of 2013 and will take several years to complete.

Peter Stroud is a principal engineering geologist and senior associate with Kleinfelder who has worked on a prominent dam removal project in the Pacific Northwest for the past several years.

Previous articleBC Hydro seeks management of work on 492-MW Bridge River system
Next articleAsset Management: Using ADCP-Based Flow Monitoring to Improve Operations at Niagara

No posts to display