Erosion of rock in spillways

Abstract

Spillways for large embankment dams typically feature an ogee structure and a section of concrete lined chute, but at some location discharge onto an unlined section. Erosion of this unlined section can require costly remediation and can potentially threaten the integrity of the dam. The unlined section of spillway is typically sited in fractured-rock. The erosion of this highly heterogeneous substance does not conform to existing erosion assessment methods for sediment transport or rip-rap. Rather, the designer simply makes a judgement, based on comparison to erosion at other dam sites. While this has been successful in many cases, there have been cases where significant erosion has occurred. In addition, flood estimates in Australia have been revised, and it is unknown how vulnerable existing spillways will be to erosion from larger forecasted floods. This dissertation was undertaken in response to this problem, and sets out to provide improved methods for assessment of erosion of unlined spillways in rock. The thesis is structured into three parts. Firstly, at the time of writing, industry relied largely on design techniques that were based on benchmarking a site in question against observed erosion at other dam sites. In Part 1 the historical development of these comparative design techniques is traced and is critically reviewed against the original geotechnical and hydraulic data, and against detailed geotechnical and hydraulic data from 29 dams in Australia, South African and the USA, which were collected as part of the dissertation. This data was used to develop rock-mass indices that specifically represent erodibility, and an improved comparative design method. In Part 2, a laboratory test program examining forces arising from water flowing over joint-bound blocks is presented. The nature of fluctuating hydraulic pressures are examined in detail, and equations allowing estimation of pressures and forces on individual rock units at the prototype scale are given. In Part 3, the findings of the laboratory studies are applied to develop analytical representation of erosion. Improved analytical representations of erosion of rockfill (rip-rap) are developed. However, it was found that a generalised analytical representation of fractured rock is not appropriate, due to the unique geology at every spillway site. A methodology for site-specific kinematic assessments is presented using the findings from the laboratory testing.