JSNH&B home • 2015 • vol. 6 no. 1

The Nuts and Bolts of Dams and Their Removal

by Jennifer Skillen

Sierra College Biology Professor

The history, politics, ethics, and emotions of the conversion of Hetch Hetchy Valley into Hetch Hetchy Reservoir are well covered elsewhere in this issue. This article will set those matters aside and focus instead on the fundamentals of what dams are, and why and how they are sometimes removed.

What is a dam?

Simply put, a dam is a structure that holds back water, thereby increasing water levels. From there it gets more complicated. Dams vary in size, from small earthen berms made by hand to monstrous structures such as the 770-foot-tall Oroville Dam (Feather River, northern California), the tallest dam in the United States. Some are made of earth, others concrete, others are rock-filled dams covered with either earth or concrete. There are gravity dams (such as Grand Coulee Dam on the Columbia River), arch dams (Glen Canyon Dam on the Colorado River), and buttress dams (Bartlett Dam on the Verde River). Dams may be considered to be storage dams, having a large amount of water behind the dam with a long residence time, or run-of-river dams, with a smaller amount of water behind the dam and a short residence time. The primary functions of a dam may include flood control, recreation, hydroelectric power generation, irrigation, and public water supply. In California the two largest uses of our reservoir water are public water-supply followed closely by irrigation (1). O’Shaughnessy Dam on the Tuolumne River is a massive concrete arch gravity dam that rises 312 feet above the riverbed, capable of impounding 444 million cubic meters of water.

How many dams are there today?

On a global scale there are certainly millions, albeit more smaller dams than larger ones. A precise number is difficult to ascertain—it depends on who you ask. The International Commission on Large Dams identified 45,000 large dams worldwide in 2000; today that number has increased to 50,000. However the US Army Corps of Engineers identified 76,500 large structures in the United States alone in 2000, and today the database maintained by the Corps, the National Inventory of Dams, lists more than 79,000 dams “that are more than 25 feet high, hold more than 50 acre-feet of water, or are considered a significant hazard if they fail.” Clearly the definition of “large” varies from agency to agency. Accounting for this variation, it is reasonable to say that there are 75,000 – 100,000 dams in the United States that are identified in various databases, with possibly as many as 2 million more small structures that are not tracked in the federal National Inventory of Dams. (1)

Where are the major dams in California?

Though most of California’s rain and snow falls in the northern part of the state, most of the people live in the southern portion. The cities, fields, and factories of the state require water, and over the past century a complex system of dams, reservoirs, canals, and pipelines has been constructed to solve the problem of transporting water, primarily from north to south.

There are two large water projects in California’s Great Central Valley, the federal Central Valley Project (CVP) operated by the US Bureau of Reclamation and the State Water Project (SWP) operated by the Department of Water Resources. Construction of the CVP began in 1937. Today it consists of 20 dams, including Shasta Dam (Sacramento River), Trinity Dam (Trinity River), Folsom Dam (American River), Friant Dam (San Joaquin River) and New Melones Dam (Stanislaus River). Construction of the SWP began in 1957 in Oroville. Today the SWP consists of “34 storage facilities, reservoirs and lakes; 20 pumping plants; 4 pumping-generating plants; 5 hydroelectric power plants; and about 701 miles of open canals and pipelines.” (2) The SWP includes Oroville Dam (Feather River), Castaic Dam (Castaic Creek and off-stream storage for the West Branch of the California Aqueduct), B.F. Sisk Dam (San Luis Creek and off-stream storage for the California Aqueduct), Pyramid Dam (Piru Creek and off-stream storage for the West Branch), and Perris Dam (off-stream storage for the East Branch of the California Aqueduct). There are also many local and regional water and power agencies that own and operate dams, such as the Nevada Irrigation District, Pacific Gas & Electric, Los Angeles Department of Water and Power, and the San Francisco Public Utilities Commission, among others.

Though in reality the demolition of a dam is a long and complex process, we can describe it simply as three steps: (1) decide if the dam should be removed; (2) if yes, determine how it should be removed; and (3) consider what happens after the dam is gone.

Why are dams removed?

The reasons for the removal of a dam, a process nearly as long and complex as the initial construction of the dam, vary but often include financial considerations and concerns over negative environmental consequences of keeping the dam. As with most man-made objects, even with constant maintenance and repair, dams eventually age and the cost of this upkeep or replacement can become prohibitive. As former Secretary of the Interior Bruce Babbitt wrote in a paper discussing dam removal, “...gradually, over the years, water evaporated from reservoir surfaces or got choked by algal blooms; concrete crumbled under pressure and time; structures severed salmon migration, collected silt, and cost millions to repair or replace.” (3) The San Clemente Dam on the Carmel River is a prime example. By 2010 this dam was no longer providing any measurable benefits— it was 90% silted-in and at risk of collapse in a major earthquake. Retrofitting the dam was estimated to be a $50 million project, while removal was $84 million. A consortium of state and federal agencies and non-profit organizations, along with the dam owner, worked together to fund the removal of the dam. The project is in-progress, with dam demolition slated for 2015. (4)

A dam is engineered and constructed to serve a particular purpose in a specific location for an estimated length of time. The primary function of a dam is to block the flow of water, but, as it does so, it also blocks the natural movement of sediment downstream through a watershed. These sediments, eroded from the upper reaches of the tributaries, are normally carried downstream, slowly settling out of the water column in the slower-moving stretches of the river, creating sand bars mid-stream and small beaches along the banks. With a dam across the river, however, that normal transport system is interrupted. The relatively deep, still waters of a reservoir allow much of the sediment to settle onto the reservoir bottom, out of the water column, before the water is gradually released through the gates of the dam. As the decades pass, more and more sediment accumulates in the reservoir, gradually reducing the amount of water that can be stored. For example, upon the completion of Matilija Dam on a tributary of the Ventura River in 1948, a report estimated that sediment accumulation would fill the reservoir in 39 years. The dam has been notched twice since then, in an attempt to allow some sediment to move downstream, but in spite of these efforts, the water storage capacity is predicted to be zero by 2020. (5)

The sediment load—the amount of sediment carried by a river—can be estimated, and a dam and reservoir constructed accordingly, to maximize the length of time the reservoir will remain functional. The average lifespan of a dam is often estimated to be 50 years. (6) Another water policy expert (7) estimates that, on average, between 0.5% and 1% of a reservoir is filled by sediment each year, meaning that most dams would have a lifespan of 100-200 years. However, the sediment load of a river can vary in unexpected ways, causing reservoirs to fill before their predicted “expiration date.” Extensive timber harvests, wildfires, or development can increase the amount of sediment entering a river.

Once sediment begins to fill a reservoir, can it be removed? Certainly dredging out the sediment is always a possibility, but it is not always cost-effective or environmentally safe. The mechanical challenges include removing sediment from the bottom of a deep reservoir—either with cranes based on floating barges, on the dam, or on shore – and then transporting that heavy, wet sediment in trucks down often narrow, winding roads to a disposal site. Finding a willing disposal site can also be difficult—and expensive—if testing shows the sediments are contaminated with heavy metals. The process of dredging—which can take months—also stirs up the sediments, making the water quality poor both in the reservoir and downstream. This can negatively impact recreation, as well as fish and wildlife populations. There are methods of minimizing the environmental impact of dredging sediments out of a reservoir, but the costs can quickly run to the hundreds of millions of dollars.

Many large dams in the United States are used to generate hydroelectric power. “All hydropower dams not owned by the federal government must obtain an operating license from FERC [Federal Energy Regulatory Commission], unless the dam has been issued an exemption or is on a non-navigable river….When those 30 to 50 year licenses expire, the dam owner must reapply to FERC to obtain a new license…As part of this licensing process, FERC must determine whether issuing a new license is in the public interest, providing equal consideration to power development and non-power uses of the river (e.g. fish and wildlife habitat, recreation, aesthetic)….” (8) This relicensing process requires documentation of the environmental impacts of dam operation and compliance with all pertinent regulations, such as the federal Clean Water Act, Clean Air Act, Endangered Species Act, etc.

Many dams that go through the FERC relicensing process obtain a new license and continue to operate, though sometimes with new operating parameters meant to better protect fish and wildlife inhabitants of the watershed (for example, the Yuba-Bear Hydroelectric Project relicensing). In some cases a new license requires the installation of fish ladders or other mitigation measures. Occasionally a renewal is denied, or the dam owner decides that it is not in their (or their shareholders) best interests to go through the relicensing process or to comply with the requirement of a new license. In these situations the dam owner and FERC may reach a mutual agreement that the dam should be removed. This is the case with four dams on the Klamath River, where the cost of relicensing outweighs the costs of removal. (9)

While it would seem obvious that O’Shaughnessy should be re-examined today, relative to its age and clear environmental impacts, the dam was authorized before the Federal Power Act. It is, therefore, exempt from FERC relicensing—removing this option from the restorationists’ toolbox.

Have any large dams been removed?

Elwha DamYes, around the world large dams are occasionally removed. However, in the United States most dams removed in the 20th century (467 partially or completely demolished) have been no more than 5 meters tall (1). Several large dams have recently been removed, including the 33 meter tall Elwha Dam and the 64 meter tall Glines Canyon Dam, both on the Elwha River in Washington. And, as mentioned earlier, the San Clemente Dam on the Carmel River is in the process of being demolished, while the removal of 4 dams on the Klamath River is in the planning stages. It is far more likely to have a small dam slated for removal than a large dam. Most dams are small, and small dams are less likely to be critical for flood control or power generation. (1) At 312 feet tall, O’Shaughnessy Dam is far larger than any other dam that has been or is planned to be demolished, though it likely has far less sediment behind it than the dams on the Elwha River, Carmel River, or Klamath River. (10)

How are dams removed?

While it is tempting to think that dam removal, once the discussions, arguments, and analyses have been concluded, ought to be a quick process dependent only on a few well-placed explosives, nothing could be further from the truth. While some dams are unintentionally removed when natural disasters strike, such as earthquakes, heavy rainfall, or mudslides, these catastrophes take a heavy toll on life and property. For example, in 2006 the earthen Ka Loko Dam in Hawaii burst after heavy rains, killing 7 people. (11) The worst such dam failure occurred in China in 1975. An unlikely combination of weather events led to a 1-in-2000 year flood. Banqiao Dam on the Ru River collapsed, followed within minutes by the failure of 62 more dams downstream. An estimated 171,000 people died during the flood and the subsequent period of famine and disease epidemics. (12) Often the effects of these natural disasters are compounded by poor construction or poor maintenance of the dams.

The controlled removal of a dam that maximizes the safety of both the human and wildlife inhabitants of the area is a complicated, slow process that is unique to each dam. For example, when planning for the removal of the two dams on the Elwha River, it was estimated that the dams were holding back a combined 21-22 million cubic yards of sediment. Rather than mechanically excavate and remove this sediment, it was determined that the best option was to gradually notch the dams, allowing the water levels to gradually decrease over two years and the sediment to erode downstream naturally. (13) Any amount of deconstruction will add sediment to the water, decreasing water quality. And clearly, the distribution of water in the river will change drastically—parts of the once full reservoir will become dry, while downstream former riverfront property will be inundated. A 1988 National Park Service report (14) describing O’Shaughnessy Dam removal alternatives spends little time on exactly how the dam would be dismantled, though it describes one alternative in which the reservoir would be lowered (drawn down) in one year, and two alternatives in which the draw down would take five years. A later study suggested that the dam would have to be dismantled by either controlled blasting, sawing, or ramming. Any of these methods will produce a great deal of debris which will have to be hauled away, possibly requiring the construction of roads to access the demolition site. (10)

In summary, there are numerous reasons for which dams are designated for removal.  When such designations are made, designs for dismantling must address costs and myriad environmental concerns. Both of these are probably far greater than when the dam was first constructed.


  1. Poff NL, Hart DD. 2002. How dams vary and why it matters for the emerging science of dam removal. BioScience 52(8): 659-668. (http://www.fws.gov/habitatconservation/Dams.pdf).
  2. DWR (Department of Water Resources). 2015. California State Water Project Overview. Accessed 1/18/15. (http://www.water.ca.gov/swp/).
  3. Babbitt B. 2002. What goes up, may come down. BioScience 52(8): 656-658.
  4. SCDR (San Clemente Dam Removal & Carmel River Re-route Project). 2015. Accessed 1/18/15 http://www.sanclementedamremoval.org/).
  5. MDERP (Matilija Dam Ecosystem Restoration Project). 2014. Facts and History. Accessed 6/2/14. (http://www.matilijadam.org/facts.htm).
  6. MIT (Massachusetts Institute of Technology). 2012. Dams. Mission 2012 Clean Water. Accessed 28 April 2014 at http://web.mit.edu/12.000/www/m2012/finalwebsite/problem/dams.shtml.
  7. Workman JG. 2007. How to fix our dam problems. Issues in Science and Technology. Fall 2007. (http://issues.org/24-1/workman/).
  8. Bowman MR. 2002. Legal perspectives on dam removal. BioScience 52(8): 739-747.
  9. American Rivers. 2009. FAQs about the Klamath Hydropower Settlement Agreement. (http://www.americanrivers.org/assets/pdfs/dam-removal-docs/klamath-hsa-faq.pdf?506914).
  10. DWR. 2006. Hetch Hetchy Restoration Study. State of California, The Resources Agency, Department of Water Resources and Department of Parks and Recreation. (http://www.water.ca.gov/pubs/environment/hetch_hetchy_restoration_study/
  11. Zimmerman M. 2014. Families still waiting for justice 8 years after Ka Loko Dam breach killed their loved ones. Hawaii Reporter. Access 1/18/15 (http://www.hawaiireporter.com/familes-still-waiting-for-justice-8-years-after-ka-loko-dam-breach-killed-their-loved-ones).
  12. Fish E. 2013. The forgotten legacy of the Banqiao Dam collapse. The Economic Observer. Accessed 1/18/15 (http://www.eeo.com.cn/ens/2013/0208/240078.shtml).
  13. Landers J. 2010. National Park Service awards construction contract to restore fish passage on Elwha River. Civil Engineering News 80(12): 14, 16.
  14. NPS (National Park Service). 1988. Alternatives for restoration of Hetch Hetchy Valley following removal of the dam and reservoir.

Images Credits

- Elwha Dam, National Park Service