Physical Hydraulic Modeling of Canal Breaches
The Bureau of Reclamation (Reclamation) has constructed more than 7,300 miles of canals since 1902. When in operation, these embankments have a 24/7 potential to breach as a result of seepage-induced erosion or overtopping flow. Although they pose threats to life and property that are similar to traditional dams, canal embankments are not regulated by the Dam Safety Program.
To quantify the potential risks, predictions are needed of breach outflow hydrographs and resulting inundation levels and flooding consequences. Technologies for simulating the breach of traditional embankment dams have some application to this problem, but there are also important differences, most of which tend to cause peak breach outflows to be lower for canals than for traditional dams. This Science and Technology (S&T) Program research project will use physical hydraulic modeling to gain a better understanding of the key physical processes and their importance in determining the resulting peak outflow. The long-term objective of the work is to develop improved numerical models for simulating canal breach events.
Need and Benefit
Failures of canal embankments on Reclamation projects have occurred with some regularity throughout our history. The last significant failure of a Reclamation canal was the Truckee Canal failure of January 2008. Before this most recent failure, the Truckee Canal had experienced as many as eight previous failures. Although no lives were lost in this case, property damage was extensive.
Loss of life during embankment failures is a very real possibility as urban development takes place in close proximity to Reclamation canals. To quantify risks presented by potential canal failures, predictions are needed of outflow hydrographs and resulting inundation and flood hazards. Tools that allow us to make these predictions will enable the prioritization of repair and rehabilitation efforts and may also allow Reclamation to identify canal reaches that should be given heightened attention, such as installing early warning systems and developing detailed emergency action plans for canal reaches that present significant threats to life.
Modeling breach outflows from canals has received little specific attention in hydraulic engineering literature. Tools for traditional embankment dams are available, but similar tools for canal breaches are lacking. A recent attempt to study the potential breach of the Hayden-Rhodes Aqueduct in Reach 12 (Scottsdale, Arizona) using the MIKE21 modeling package highlighted the fact that some unique aspects of a canal breach are still poorly understood, requiring significant assumptions to be made to complete the study.
While numerical models can compute accurate estimates of the inundation conditions resulting from a breach of a particular size, the prediction of the breach size itself is still filled with uncertainty. In addition, the design of numerical and physical models requires the delineation of a physical space to be included in the model and the definition of boundary conditions at the edge of that space. The boundary conditions used in the Hayden-Rhodes Aqueduct simulation were probably too limiting and shortened the duration of the outflow hydrograph. Also, the simulation did not allow for any progressive erosion and enlargement of the breach (breach size was chosen as an input). Boundary conditions of the model would have also limited the extent of erosion, if erosion had been included.
The proposed physical model tests carried out in this project would improve our understanding of the canal breaching process and facilitate the development of better numerical models for future use. We also plan to work to provide an array of new tools for simulating canal breach events, some targeted at quick appraisal-level evaluations and others focused on more detailed studies of high-hazard canal reaches.
Lower Colorado Regional Office, Lower Colorado Region
Contact the Principal Investigator for information about these documents.
This information was last updated on May 22, 2013
Contact the Research and Development Office with questions or comments about this page