There is a large stock of embankment dams throughout the world needing the assessment of their
safety as required by modern dam safety regulations. Due mainly to economic and site constraints
associated with potential dam upgrading work, it is imperative that a rational approach be adopted in
assessing their safety and in designing the remedial works. One of the most important criteria is the
selection of appropriate geotechnical parameters under different conditions. Predominant loading
conditions in a dam are much different from those in other structures such as bridge and building
foundations and therefore the direct adoption of traditional approaches may not always be valid. This
paper presents the various aspects of issues associated with the stability assessment of dams including
the rational selection of the parameters and numerical codes available to dan/geotechnical engineers
to assess their safety.
Mike Taylor, Jonathan Jensen and Greg Branson
Pykes Creek Dam is a 33 m high, 22,120 ML embankment dam, 72 km west of Melbourne owned and operated by Southern Rural Water.
The outlet works include a 30 m high “wet” outlet tower near the upstream toe of the dam on the right
abutment with its lower half comprising a concrete lined shaft excavated in rock. A 1.5 m diameter
concrete lined tunnel extends 30 m upstream from the base of the tower to a reinforced concrete inlet structure.
The only controls upstream of the downstream toe of the dam comprised 2 guard gates located on the downstream side of the tower, operated manually by means of handwheels from the top of the tower.
Major deficiencies with the outlet works included:
A major constraint in addressing these deficiencies was that any remedial works needed to be
undertaken without draining the reservoir or interfering with the releases required for downstream
consumers, including irrigators in Werribee and Bacchus Marsh.The paper describes how all of the deficiencies have been addressed with no interruption to supply, by means of a collaborative effort between the dam owner, the consulting engineer, and 5 separate contractors, with the dam owner playing a leading role.
Construction of the Lake Buffalo Dam was completed in 1965. It was to be a temporary dam, required to operate for several years, then act as a cofferdam for the construction of a much larger dam downstream. This larger dam was never built and a risk assessment completed by Goulburn Murray Water (G-MW) in 2001 identified several dam safety deficiencies at Lake Buffalo were among the highest priorities for risk reduction measures across the G-MW dams portfolio. Specifically it identified Lake Buffalo as having inadequate flood capacity and there were also concerns about transverse cracking within the embankment.
This paper describes the detailed investigation and analysis of the embankment cracking including assessing the potential for piping through an embankment having deficient filters and known transverse cracking. The design features of the upgrade are also described including the design of the a filter buttress, a parapet wall raise, Computational Fluid Dynamics (CFD) modelling and spillway anchoring. Construction was completed in 2003.
Barton Maher, Richard Rodd
Changes to the estimation of extreme rainfall events resulted in significant increases in the estimates of the PMF since the original design of Wivenhoe Dam. To upgrade the dam to meet these new requirements, SEQWater (owner and operator) formed an Alliance with Leighton Contractors, Coffey Geosciences, MWH and the NSW Department of Commerce.
The option selected for the upgrade works included the construction of a new secondary spillway, upgrade of the existing gravity section, radial-gated spillway, and strengthening of the dam crest.
Value management was key throughout the project ensuring the Alliance was continually looking to
improve practices, increase cost-effectiveness and create innovative solutions for design elements of the project.
On numerous occasions when the design was challenged, the Alliance made ‘best for project’ decisions to carry out additional investigations or design work to pursue alternatives. As an example, the powerful tool of Computational Fluid Dynamics was used in the analysis and design of flow deflector plates on the existing spillway, which were an alternative to the originally designed gate locking pins. The investigation and development of this alternative resulted in significant cost savings and a more effective design solution.
This paper presents aspects of the design carried out by the Wivenhoe Alliance, lessons learned, and the way continual investigations during construction provided value for money solutions.
Paul Hurst, Michael Smith
Wellington Dam is an extreme hazard concrete gravity dam located on the Collie River approximately 170km south of Perth. Originally constructed to a height of 19m in 1933, the dam was raised to its present height of 34m in 1960 by placing significant additional concrete against the downstream face of the original dam. To ensure a lasting bond along the interface between the original and secondary concrete, an open slot was formed and later grouted once the temperature of the secondary concrete was similar to that of the original dam.
A recently completed stability analysis identified that Wellington Dam falls well short of contemporary dam engineering standards for flood loading. Several assumptions were made during the preliminary analysis relating to concrete shear strength parameters, bonding between the original and secondary concrete and drain effectiveness that generated a significant range of results. On this basis, further investigation was carried out to define the concrete parameters and drain condition at Wellington Dam.
Exploratory drilling found that Wellington Dam is cracked from the upper gallery through to the downstream face. The drilling programme also confirmed that the interface between the original and secondary concrete has become unbonded and that the gravity dam is behaving like an unbonded short composite beam. The mechanism causing the observed behaviour of Wellington Dam can largely be explained by external temperature effects and Alkali Aggregate Reaction, (AAR).
This paper explores the techniques used to investigate the condition of the concrete and illustrates the relationship between concrete behaviour and temperature and AAR effects within a composite concrete gravity dam
The Wivenhoe Dam Spillway Augmentation Project involved the construction of an additional spillway on the right abutment of the main dam. The right abutment is located in massive sandstones and siltstones of Jurassic and Upper Triassic age.
Seismic refraction surveys and borehole drilling conducted at the design stage for the project
indicated that part of the spillway area was likely to be marginally rippable to unrippable using a
Caterpillar D9 bulldozer or equivalent. Further assessment and rock strength testing was conducted during the initial stages of excavation where D9 and D10 bulldozers were in operation. The results from this further work indicated that a section of the spillway extending from the proposed position of the ogee crest to approximately 100m further upstream were unlikely to be unrippable for a D9 dozer and marginally rippable for a D10.
Excavation options considered for this section included full scale blasting and load out, limited small scale ‘popping’ combined with ripping or the use of larger ripping equipment. Based on an
assessment of cost-benefit, and given the availability of larger ripping equipment, it was decided to
use a combination of D10 dozers and a Komatsu 475A bulldozer (D11 equivalent) equipped with
single tine ripping tools. The use of this equipment proved successful with better than anticipated
production rates being achieved. This resulted in significant cost and time savings for the project and reduced the likelihood of potential adverse impacts on the existing dam grout curtain, environment, travelling public and residents that may have occurred during blasting.