Skip to main content
ARS Home » Research » Publications at this Location » Publication #167123

Title: EARTHEN SPILLWAYS DESIGN AND ANALYSIS - STATE OF PRACTICE

Author
item Temple, Darrel

Submitted to: Workshop Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 8/1/2003
Publication Date: 7/1/2004
Citation: Temple, D.M. 2004. Earthen spillways design and analysis - state of practice [abstract]. In: Proceedings of the FEMA Workshop. Issues, Remedies, and Research Needs Related to Dam Service and/or Emergency Spillways, August 26-27, 2003, Denver, CO. 2004 CDROM.

Interpretive Summary:

Technical Abstract: Earth channels have been widely used for auxiliary or emergency spillways to convey major flood flows around dams. These spillways are normally designed to flow infrequently and are generally considered have operated successfully if erosion experienced during a given event does not threaten the integrity of the dam and reservoir. For watershed flood control reservoirs, the typical earth spillway consists of a vegetated channel with an inlet reach, a level crest section, and one or more exit channel reaches designed to flow supercritical at design discharge. Larger structures may incorporate concrete weirs or sills to provide improved hydraulic control characteristics, including more uniform flow conditions over the width of the spillway. Energy dissipaters and erosion barriers of various forms may also be integrated into the design, but this discussion focuses on the behavior of the earth spillway channel itself. Historically, earth or vegetated earth spillways designs were based on stable channel design criteria described in publications such as the USACE "Hydraulic Design of Flood Control Channels" or on an empirical bulk length as described in USDA TR-52. Although failure of these spillways has been rare, erosion observed during spillway flows has led to refinement of design and analysis procedures in recent years. The United States Society on Dams is presently developing a bulletin describing the history and the present state of the science in the area of erosion of earth spillways in more detail. Publication of this bulletin is expected during 2004. The methods presently being used for the design and analysis of earth spillways tend to be semi-empirical, based on flows and erosion observed during the past 25 years. The REMR erosion prediction method developed by the US Army Corps of Engineers consists of a classification system that allows comparison of an erosion risk class with an erosion potential class. The approach predicts whether erosion is or is not expected. The vegetated earth spillway erosion model developed by USDA and incorporated into the NRCS Sites software divides the erosion process into three sequential phases: 1) failure of the vegetal cover, if any, and development of concentrated flow; 2) surface detachment in the area of concentrated flow leading to development of a vertical or near vertical headcut; and 3) deepening and upstream advance of the headcut. Each phase of the process is described by different threshold-rate relations reflecting the physics of that phase. The model is applied iteratively to various potential points of initiation to determine the scenario with the greatest potential for spillway breach. The USDA model represents a first attempt at quantification of the dominant spillway erosion processes. The potential exists for refinement of the relations describing all phases of the overall process. The US Army Corps of Engineers has applied the general approach with modified equations and ongoing research is expected to result in improved relations describing the processes. Research is also underway to refine the three-phase approach for application to prediction of dam breach due to overtopping flows. There is a need for continued study to develop improved parameters for describing the resistance of geologic materials to erosion and for improved means of identifying pertinent characteristics of materials that may be exposed during the erosion process. Other identified research needs include expanding the current breach prediction model to include spillway erosion that occurs after the initial breach and identification of conditions where modes of failure other than headcut formation and advance dominate the process.