Skip to main content
ARS Home » Research » Publications at this Location » Publication #113078

Title: BREACH MORPHOLOGY OBSERVATIONS OF EMBANKMENT OVERTOPPING TESTS

Author
item HAHN, WILLIAM - OKLAHOMA STATE UNIVERSITY
item Hanson, Gregory
item Cook, Kevin

Submitted to: American Society of Civil Engineers Water Resources Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 8/2/2000
Publication Date: 8/2/2000
Citation: Hahn, W., Hanson, G.J., Cook, K.R. 2000. Breach morphology observations of embankment overtopping tests. American Society of Civil Engineers Water Resources Conference Proceedings, Minneapolis, MN.

Interpretive Summary: Even though dam failures are rare, the impact to people and property downstream of a reservoir is important to engineers and planners alike, who must evaluate the hazards of local flooding. Recent outdoor laboratory studies were conducted to simulate dam failures due to floodwaters running over the top of an earthen embankment. Overtopping tests were conducted in an outdoor flume and in relatively large-scale models with various soil materials to observe impact of different materials on the type and rate of embankment failures. Testing observations showed the erosion process had many phases. In the initial phase the downstream surface of embankment eroded in a stair-stepped fashion of small multiple waterfalls. In the next phase these multiple waterfalls ultimately merged into a single upstream-migrating waterfall. In the final phases the waterfall migrated upstream far enough to begin lowering the embankment elevation and the reservoir storage pool began to drain. Rates of erosion damage to the embankment were observed to vary as much as 60 times depending on the material tested. This study adds insight into the nature of embankment overtopping erosion and failure. These observations will be essential in updating earthen embankment failure models.

Technical Abstract: Dam and embankment breaching from overtopping events is important to both engineers and planners alike, who must predict impacts on local communities and surrounding areas affected by flooding. Most present breach prediction models, however, have assumed more simplistic breach morphologies than those observed during actual breach events. Recent research on overtopping events from both an outdoor flume and full-scale models, with soils that ranged from non-plastic SM silty sand to CL lean clay, have shown that the erosion process of embankments has a significant impact on timing and rate of breach. This indicates that the models using simplistic breach morphology analysis may be inconsistent with the process actually occurring. Testing observations showed the erosion process was multi-phase: initial downstream surface erosion progressed into stair-stepped multiple overfalls, which ultimately merged into a single upstream-migrating headcut. Simultaneous erosion processes included gully and breach widening and upstream headcut migration, the rate of which was a function of embankment material properties. Rates of migration were observed to vary as much as 60 fold as materials for the different soil materials tested. This research adds insight into the nature of breach morphology, including formation characteristics and criteria, general morphological development, and rate of change during overtopping events. Flood prediction models need to integrate updated breach, hydraulic, and material parameters in order to generate more realistic flood hydrographs, and to ensure appropriate floodwater routing.