Africa, Asia and Oceania, Dam Safety, Environmental, Europe, Latin America, North America

Dealing with Sediment: Effects on Dams and Hydropower Generation

Issue 1 and Volume 25.

How do reservoir sedimentation and appropriate management techniques affect operations of dams and hydroelectric facilities? The authors cover the topic and provide illustrative case studies, including the 2,100 MW Aswan High Dam in Egypt.

By Greg Schellenberg, C. Richard Donnelly, Charles Holder and Rajib Ahsan

Although sedimentation of the world’s reservoirs represents a serious threat to the sustainability of hydropower, there is limited guidance on how best to address the problem. Sedimentation affects the safety of dams and reduces energy production, storage, discharge capacity and flood attenuation capabilities. It increases loads on the dam and gates, damages mechanical equipment and creates a wide range of environmental impacts. This article explores sedimentation issues as they pertain to hydropower facilities, dam safety and the environment; discusses sedimentation management techniques; and describes how they can be implemented to limit the impacts on hydropower.

Background

Reservoir sedimentation is a process of erosion, entrainment, transportation, deposition and compaction of sediment carried into reservoirs formed and contained by dams. In unregulated, mature rivers with stable catchments, sediment processes are relatively balanced. Construction of a dam decreases flow velocities, initiating or accelerating sedimentation,1 resulting in progressively finer materials being deposited (see Figure 1).

There are three stages in a reservoir’s life:2

  • Continuous and rapidly occurring sediment accumulation;
  • Partial sediment balance, where often fine sediments reach a balance but coarse sediments continue to accumulate; and
  • Full sediment balance, with sediment inflow and outflow equal for all particle sizes.

Most of the world’s reservoirs are in the continuous accumulation stage.2 Many were designed by estimating sedimentation rates in order to provide a pool with sufficient volume to achieve a specified design life. However, this design life is typically far less than what is actually achievable. Therefore, managing reservoirs to achieve a full sediment balance is essential in order to maximize their lives.

Developing regions of the world that stand to benefit most from hydroelectricity are often those with the highest sediment yields (see Figure 2 on page 8).3 In these regions, sustainable hydropower development must involve consideration of sediment management techniques during design, construction and operation.

Sediment impacts on generation

About 0.5% to 1% of the total volume of 6,800 km3 of water stored in reservoirs around the world is lost annually as a result of sedimentation.2 As a result, global per capita reservoir storage has rapidly decreased since its peak at about 1980. Current storage is equivalent to levels that existed nearly 60 years ago.2

Loss of reservoir storage reduces flexibility in generation and affects the reliability of water supply. Without storage, hydropower facilities are entirely dependent on seasonal flows. These flows might not occur when energy is needed, eliminating one of the key benefits that hydropower provides over other renewables.

Sediments discharged from an upstream dam in a cascade system can increase tailwater levels, reducing power generation.1 This would impact the generation potential of all plants in the cascade and increase the possibility of powerhouse flooding.

Sediment impacts on stability

Sediment loads are commonly idealized as a static at-rest soil pressure. The U.S. Bureau of Reclamation’s design manual for small dams suggests that sediments be considered equivalent to a fluid with an implied pressure coefficient of about 0.39 and an internal friction coefficient of about 37 degrees.

However, actual reservoir sediment properties can vary considerably. Unconsolidated fine-grained sediments likely have lower shear resistance and a higher at-rest pressure coefficient, while a reservoir filled with coarser sediments may have a higher shear strength.1

Published criteria with respect to potential changes in uplift pressures due to sedimentation often neglect the fact that fine-grained sediments may reduce uplift in the same manner as does an engineered upstream blanket. Conversely, in the case where there is a large turbid inflow, higher uplift pressures would be expected until enough particles had settled to form a blanket.

During a seismic event, it is likely that liquefied sediments would quickly return close to their original state, resulting in a rapid dissipation of pore pressures. Therefore, it may be questionable to automatically assign higher uplift pressures in this case.

Commonly used design considerations can omit some plausible load cases. For example, an underwater sediment slope failure could cause surface waves, adding additional loading, hydro-dynamic pressure waves and an inertial loading from the dense fluidized soil-water mass. Another phenomena commonly ignored relates to turbidity currents in reservoirs. Such turbid fluid with a sediment load of 100 mg/l could be about 6% heavier than clear water.1

Submarine landslides are widely studied because of their potential to create tsunami waves. However, designers also need to consider the potential that failure of the steeply sloped deltaic front could increase loading and produce compression waves that may fluidize finer sediments near the toe of the landslide. As the deposition advances toward the dam, the potential for issues progressively increases.

It is often assumed that, during an earthquake, sediments fully liquefy, lose all strength and exert a dense fluid hydrostatic load on the dam. However, this degree of fluidization likely is not possible in a reservoir filled with coarse materials. Designers also often assume that the fully fluidized dense fluid contributes to hydro-dynamic pressure loading based on Westergaard’s formula, ignoring the physical basis for its derivation. In fact, there is some question about the applicability of Westergaard’s formula for hydro-dynamic pressures.

Designs also need to consider the degree of saturation of the sediments. There is minimal system damping under dynamic loading when reservoir sediments are fully saturated. However, significant reductions in acceleration occur when sediments are partially saturated.4,5 For rigid foundations, hydrodynamic pressures decrease slightly at the dam base when sediments are fully saturated but increase when partially saturated.4 Partial saturation will increase he system’s response to horizontal ground movement.5 Sediment thickness is an important consideration, especially when the sediments are partially saturated.5 Thin layers result in minimal absorption of horizontal motions, largely due to a relatively high modulus of elasticity and low attenuation coefficient.6 Over the reservoir life cycle, this changes as sediments continue to accumulate.7 Other important factors are sediment density, compressibility and pore water pressure.5,7

This dependence on sediment properties makes a strong case for their measurement and inclusion as part of the design.4 However, designs are performed before sedimentation occurs and the same sediments that are stable under normal conditions and absorb energy at the bottom of the reservoir could liquefy. For this reason, the use of a reservoir bottom reflection coefficient must be logically linked to assessment of the reservoir sediment behavior and ongoing monitoring.

Sediment impacts on discharge capability

Sediments will often block low-level outlets designed to allow for reservoir drawdown.1 As sedimentation continues, clogging of spillway tunnels or other conduits may occur.1

Reduction of spillway capacity can occur as a result of the loss of approach depth when the sediment front reaches the dam. The reservoir becomes a delta-filled valley that takes a meandering course such that a flood wave does not spread out to allow flood routing.

Sediment impacts on equipment

Sediment can damage turbines and other mechanical equipment through erosion of the oxide coating on the blades, leading to surface irregularities and more serious material damage.8 Sustained erosion can lead to extended shutdown time for maintenance or replacement.8

Many factors determine rates of mechanical abrasion. Of particular importance is sediment type and physical characteristics. Angular sediments composed of minerals with a Mohs hardness greater than 5 – such as quartz, feldspar and tourmaline – are problematic. In addition, hydraulic and facility operation parameters such as flow velocity, hydraulic head, turbulence, turbine rotation speed and turbine material affect abrasion susceptibility. Impulse turbines, such as Pelton or Turgo, are more susceptible to abrasion than are reaction turbines.8 However, runner changes and needle tip/seat ring replacement are much easier with Pelton turbines. Therefore, they may be preferable on the basis of the overall life cycle cost.

Abrasion can be reduced by selecting metals to increase erosive resistance and/or by reducing the volume of fine sediment that reaches mechanical equipment. Plants often are designed to remove most of the coarse sediment particles. However, even silt can cause significant abrasion if the quartz content and pressure head is high enough.9 The 1,500 MW Nathpa Jhakri hydroelectric plant in India used four desilting chambers that were successful in removing coarser sediments. However, damage from the finer particles was so severe that parts of the turbines had to be replaced within one year.

Materials used commonly in sediment-prone hydropower plants are stainless steels that are heat treated for hardening and increased protection from abrasion.8 Protecting mechanical equipment from sediment abrasion can also be achieved with hard surface coatings of ceramic paints or pastes or with hard facing alloys.8 Research has shown improved resistance to sediment abrasion when tungsten carbide-based composites are used as a surface coating.8 In undertaking such assessments, it is important to consider the fact that abrasion will increase as the reservoir fills. The Nozaki method can be used to assess turbine repair frequency. The method accounts for the effective sediment concentration, particle size and shape, the turbine material and any coatings.

Turbine designs need to minimize peak velocities to reduce impacts. For a Pelton turbine, fewer jets and larger runner buckets with larger radii reduce centrifugal forces between the sediment and runner surfaces. Regardless of the turbine selected, designs must consider issues such as the ease of runner removal for future maintenance.

Sediment impacts on the environment

Any dam will cause some degree of sediment starvation downstream. Plant and animal species are sensitive to alteration of both the sediment supply and flow regime.2,10 Increases in sediment concentration can create turbid waters with a smaller euphotic zone. This decreases plant productivity, negatively impacting fish and bird species2 and causing abrasion of fish gills, thus increasing potential for disease or mortality. Turbidity can also cause visual impairment for predatory fish, affecting their feeding habits. Finally, sediment is a primary carrier of suspended pollutants such as nitrogen, phosphorous and heavy metals.10

Sediments released as a result of sediment management or a dam breach may have environmental effects that can persist for decades.

Numerical modeling of sedimentation and sediment management strategies

A variety of tools are available for hydromorphological simulation in order to optimize reservoir management:

  • U.S. Army Corps of Engineers’ HEC-RAS model features a movable boundary sediment transport calculation module that was recently used to simulate sedimentation processes resulting from hydropower development in northern Manitoba.11
  • MIKE 21 is a two-dimensional hydrodynamic model used to simulate sedimentation processes that was used to assess sediment deposition patterns and simulate the results of future flushing operations at Boegoeberg Dam in South Africa.12
  • The hydrodynamic, sediment transport and physical habitat model FAST is used to simulate morphological processes and changes to fish habitat within alluvial rivers.13 It was used to predict hydromorphological conditions and to optimize sediment flushing procedures prior to constructing new hydropower facilities on the Nile River.

Sediment management solutions

Developing and retaining sustainable storage to satisfy global needs requires inclusion of reservoir sediment management practices at project conception and throughout its life cycle. These practices vary depending on the tope of facility. For run-of-river projects, sediment management aims to remove sediments that can cause abrasion of the turbines and clog cooling water intakes. In a storage project, this objective and extending reservoir longevity are key.

For storage hydro, sediment management strategies to extend reservoir longevity can be classified into three categories:

  • Those that divert some of the sediment through or around the reservoir;
  • Those that remove or rearrange sediment that has already been deposited; and
  • Those that minimize the amount of sediment reaching the reservoir from upstream.

Many dam operators have implemented sediment management techniques designed to achieve these goals.14 Some examples are described below.

Bypassing

On-stream sediment bypassing diverts part of the sediment-laden water around the reservoir, typically using a weir that operates during high flows when sediment concentrations are high.

An off-stream reservoir can be used such that only the clear water is diverted over a bypass weir. An off-stream reservoir typically has limited capacity and can only exclude sediments carried by higher streamflows.2 However, it does reduce the amount of suspended sediment and bedload reaching the reservoir.14 Other advantages include the fact that the reservoir and dam are located away from the main river channel, allowing for minimal disruption to aquatic species and habitat and reducing the need for large on-stream spillways.2 On the other hand, off-stream reservoirs typically do not permit maximization of generation capacity, especially in areas that depend on high stream flows occurring over a short period of time.2

Sediment bypassing works best in areas of high relief where the sediment-laden flows are carried efficiently through the diversion tunnel or channel. Bypassing is most cost-effective at dams that are on the bend of a river, as this allows for a relatively short diversion between the weir and the downstream side of the dam.14

Sluicing/drawdown routing

This technique involves lowering the reservoir water levels in advance of high streamflows so that water and sediment can be routed through the spillway at high velocities. Refill occurs during the receding limb of the flood hydrograph.2,14 Sluicing methods depend on the facility’s hydrologic characteristics and reservoir size.

Dredging

Dredging can be efficient but it will continue for the life of the project and can have significant cost impacts. For example, dredging of 6 million m3 of sediment at the Loiza reservoir in Puerto Rico in 1997 cost $10/m3.1,2

Flushing

Flushing involves emptying the reservoir by opening bottom outlets and allowing the incoming streamflow to scour sediment.2,14 The effectiveness varies but, generally, only a “core” of sediment along the original channel thalweg is flushed. Sediments on the sides of the reservoir remain in place.2

An alternative method is pressure flushing, where the reservoir is partially drawn down before flushing. This redistributes coarse upstream sediments closer to the dam, alleviating their impacts, but often does not clear the finer sediments.2 Pressure flushing is also used for sediment redistribution, moving them to a less sensitive location.

Erosion control

Many watersheds experience increased erosion rates due to land use and other human practices. Erosion reduction techniques fall into three categories: structural or mechanical, vegetative and operational.1

Structural or mechanical measures – such as terraces, conveyance channels, check dams and sediment traps1,14 – decrease overland or channelized flow velocity, increasing surface storage and thereby reducing the sediment load in the runoff.

Vegetative erosion control takes advantage of plants’ natural ability to limit erosion. Agricultural practices that minimize sediment yield are particularly effective.

Operational measures minimize erosion through planning, management and organization. Examples include timing construction work such that erosion is minimized or scheduling timber harvesting to coincide with favorable soil conditions.1

Erosion management is perhaps the most widely recommended but most poorly implemented sediment management technique because land users may not see any direct benefits from controlling sediment yield.2

Selection of optimal sedimentation management techniques

The appropriate sedimentation management practice is a function of the reservoir life, expressed as the ratio of reservoir volume (CAP) to mean annual sediment inflow to the reservoir (MAS), and retention time represented as a function of the ratio of CAP to the mean annual incoming flow (MAF). Selection of the optimal sediment management techniques can be estimated based on precedent experience and these factors (see Figure 3).

Case studies

The case studies discussed below illustrate a range of sediment management concerns as well as strategies taken to mitigate them.

Aswan High Dam, Egypt

The 2,100- MW Aswan High Dam project on the Nile River in Egypt includes a 111 m high dam that impounds a 130 km3 reservoir.15 This dam has been controversial, largely due to concerns regarding sediment starvation of the Nile River Delta.15

Before construction of this dam, the Nile River transported an average of 100 x 106 tons/yr of sediment to the Nile River Delta in the Mediterranean Sea.16 Today, with a trapping efficiency of 99%, little sediment reaches the delta.15,16 While the live storage capacity of the Lake Nasser/Nubia reservoir upstream of Aswan High Dam is not expected to be compromised for another 300 to 400 years,17 the adverse downstream impacts have been widely reported.15 Erosion along the Mediterranean coast of Egypt has been ongoing for centuries, but the sediment trapping has combined with sea-level rise and other factors to exacerbate coastal erosion problems.15

Dez Dam, Iran

The 520 MW Dez hydroelectric project in southwestern Iran features a 203 m-high concrete arch dam. Reservoir sedimentation has caused the riverbed to rise by about 2 m per year, resulting in the loss of about 19% of reservoir storage during its 40 years of operation. As of 2016, the reservoir bed was now within 12 m of the power intakes, such that sediment may be drawn into the tunnels within a decade.

Sediment management strategies considered for the Dez project included watershed management, sediment flushing, tactical dredging near the power intakes, and heightening the dam. The optimal solution was determined to be sediment flushing, managed by means of powerhouse and spillway operation changes.

Another issue was the fact that sediments had risen above the low level outlets. As sluicing of the sediments through the Howell-Bunger valves introduced a risk of damage to the valves, a physical model was built to evaluate replacing these valves with radial sluice gates. Results showed that the downstream river reach could not tolerate the amount of scour associated with this modification, so the Howell-Bunger valves were redesigned with abrasion-resistant materials.

Three Gorges, China

In China, the extent of this issue has led to the development of innovations in sediment management.14 Four main sediment management strategies have been adopted. They are: storing the clear and releasing the turbid, releasing turbidity currents, sediment flushing, and dredging.19

The 22,500 MW Three Gorges Project on the Yangtze River is the world’s largest hydropower facility. The dead storage portion of the Three Gorges reservoir (17 billion m3) is designed to be filled with sediment in about 120 to 150 years. The remaining 22 billion m3 is to be retained indefinitely by flushing.19 During the June to September flood season, when 50% to 60% of the annual runoff transports much of the sediments in Chinese rivers,19 operators draw down the reservoir, retaining clearer water for the rest of the year. This strategy has been effective for reducing sediment impacts at both Three Gorges Dam and the Sanmenxia Reservoir, with a 400 MW powerhouse.19

Conclusions

The world’s reservoirs are used for many purposes, among them to provide reliable water supply, hydropower and flood mitigation. Sustainable hydropower requires dealing with the important issue of reservoir sedimentation.

This article describes sedimentation processes, identifies key impacts of sedimentation on hydropower facilities and presents techniques that can be used to address these impacts. Sedimentation can affect hydropower production due to loss of reservoir storage and/or damage to the facility’s mechanical components. Sediments deposited in reservoirs may affect the safety of dams and, without proper management, negatively impact the environment.

Methods of managing sediment fall under three general categories: those that divert sediment around or through the reservoir, those that remove deposited sediments, and those that minimize the amount of sediment reaching the facility in the first place. A variety of sediment management strategies have been used around the world, with many successful implementations documented.

This discussion highlights the need for appropriate sediment management at hydropower facilities and shows how this can be achieved through consideration of sediment concerns from the earliest design phase through to construction and operation.

Editor’s Note: This is an abridged version of a much longer article on the topic. To read the article in its entirety, visit www.hydroworld.com/index/hydro-library.html.

Notes

1. Reservoir Sedimentation Manual, McGraw-Hill, New York, 1998.

2. Morris, G., G. Annandale and R. Hotchkiss, “Reservoir Sedimentation,” in Sedimentation Engineering: Processes, Measurements, Modeling, and Practice, American Society of Civil Engineers, Reston, Va., U.S., 2008.

3. Grummer, J., “Combating Silt Erosion in Hydraulic Turbines,” HRW-Hydro Review Worldwide, Volume 17, No. 1, March 2009.

4. Bougacha, S., and J. Tassoulas, “Seismic Response of Gravity Dams II: Effects of Sediments,” Journal of Engineering Mechanics, Volume 117, No. 8, 1991, pages 1839-1850.

5. Dominguez, J., R. Gallego, and B. Japon, “Effects of Porous Sediments on Seismic Response of Gravity Dams,” Journal of Engineering Mechanics, Volume 117, No. 8, 1997, pages 302-311.

6. Hatami, K., “Effect of Reservoir Bottom on Earthquake Response of Concrete Dams,” Soil Dynamics and Earthquake Engineering, Volume 16, 1997, pages 407-415.

7. Gogoi, I., and D. Maity, “Influence of Sediment Layers on Dynamic Behavior of Aged Concrete Dams,” Journal of Engineering Mechanics, Volume 133, No. 4, 2007, pages 400-413.

8. Dorij, U., and R. Ghomaschi, “Hydro Turbine Failure Mechanisms: An Overview,” Engineering Failure Analysis, Volume 44, 2014, pages 136-147.

9. Annandale, G., G. Morris, and P. Karki, “Sediment Management at Reservoirs and Hydropower Plants: World Bank Technical Note,” Proceedings of 84th ICOLD Meeting, International Commission on Large Dams, Paris, 2016.

10. Ahmari, H., et al, “Assessment of Erosion and Sedimentation for Hydropower Projects on the Lower Nelson River, Manitoba,” Canadian Dam Association, Montreal, Quebec, Canada, 2013.

11. Kenny, S., et al, “Assessment of Impacts to the Sedimentation Environment for the Keeyask Generating Station Project using Numerical Modelling,” Proceedings of 2014 Canadian Dam Association Conference, Canadian Dam Association, Montreal, Quebec, Canada, 2014.

12. Sawadogo, O., and G. Basson, “2D Hydrodynamic Modelling of Sediment Deposition Processes and Flushing Operation of Boegoeberg Dam, South Africa,” Proceedings of 2016 ICOLD Conference, International Commission on Large Dams, Paris, 2016.

13. Bui, M., and P. Rutschmann, “Numerical Modelling for Reservoir Sedimentation Management,” Proceedings of Sixth International Conference on Water Resources and Hydropower Development in Asia, 2016.

14. Kondolf, G., et al, “Sustainable Sediment Management in Reservoirs and Regulated Rivers: Experiences from Five Continents,” Earth’s Future, Volume 2, 2014, pages 256-280.

15. Abd-El Monsef, H., S. Smith, and K. Darwish, “Impacts of the Aswan High Dam After 50 Years,” Water Resources Management, Volume 29, 2015, pages 1873-1885.

16. Milliman, J., and R. Meade, R. “World-wide Delivery of River Sediment to the Oceans,” Journal of Geology, Volume 91, No. 1, 1983, pages 1-21.

17. Smith, S., “A Revised Estimate of the Life Span of Lake Nasser,” Environmental Geology and Water Sciences, Volume 15, 1990, pages 123-129.

18. Steele, R., et al, “Sedimentation Issues in the Dez Dam Reservoir, Hatch H.G. Acres Conference, 2006.

19. Wang, Z., and C. Hu, “Strategies for Managing Reservoir Sedimentation,” International Journal of Sediment Research, Volume 24, 2009, pages 369-384.


Greg Schellenberg is junior hydrotechnical engineer, C. Richard Donnelly is principal consultant, Charles Holder is senior civil consultant, and Rajib Ahsan is senior hydraulic engineer with Hatch.