Phillip Jordan, Alan Seed, Rory Nathan, Peter Hill, Eva Kordomenidi, Clive Pierce, Michael Leonard
This paper discusses the stochastic framework that was used to generate the 5449 sets of inflow hydrographs, to develop and stress test a dam operations model. The stochastic simulations were driven by 600 different space-time patterns of rainfall generated using a stochastic space-time multiplicative cascade model. Eight significant storms were identified in the radar archive to identify parameter sets for the stochastic generation algorithm and 600 replicates of space-time rainfall were generated. The statistical properties of spatial patterns of 48-hour rainfall bursts on eight major subcatchments of the Brisbane River catchment from the 600 stochastic replicates were verified against the same statistics derived from 38 major flood causing rainfall events observed in the catchment. The hydrographs were generated using an URBS rainfall runoff routing model of the Brisbane River catchment, which was calibrated to 38 historical flood events (between 1955 and 2013) and tested on a further 10 historical flood events (between 1887 and 1947).
The stochastically simulated sets of inflow hydrographs were then used to assess the impact of variations in flood operation rules for Wivenhoe and Somerset dams. The stochastically generated events exhibit substantial variability in runoff hydrographs but with variability that is statistically consistent with observed events. The stochastically generated hydrographs provide a considerably more realistic basis for testing the outcomes for different flood operations strategies than the single design event approaches that have previously been adopted.
Francisco Lopez and Michael McKay
At 36 m high and completed in 1902, Barossa Dam is one of the first true concrete arch dams in the world. During the 1954 Darlington Earthquake the dam sustained some damage, in the form of several vertical cracks on both dam’s abutments. In 2013, GHD conducted a nonlinear time-history seismic assessment of Barossa Dam. The analyses, carried out using finite element techniques, included ground motion loading corresponding to Maximum Design Earthquakes (MDEs) with 1 in 10,000 Annual Exceedance Probability (AEP).
This paper will explain the purpose of the study, the material investigation phase, the methodology, model results, the anticipated seismic behaviour of the dam wall, as well as the predicted level of damage under the MDEs. The paper examines the dam construction practices of the beginning of the 20th century, and how such practices affected the material properties and the structural performance of Barossa Dam.
Two techniques were used to calculate seismic hazard at a number of locations in southeast Australia. To simplify matters only Peak Ground Accelerations were compared.
The first technique used a seismological model of areal source zones that was based on the recorded seismicity as well as geological and tectonic inputs. Each zone was assigned a rate of earthquake activity that had been calculated from the recorded seismicity and a magnitude completeness function. Known geological faults that are also part of the model had to be excluded to allow a direct comparison with the second technique. A standard probabilistic seismic hazard analysis then gave PGA values versus return periods. This is the approach that has been used for the current Australian earthquake loading code (AS1170.4).
The second technique used a simple historical approach whereby recorded earthquakes were combined with an attenuation function to directly give the estimated return periods. This approach takes no account of tectonics, geological terranes or faulting – it simply uses the known, recorded earthquake catalogue. This is the technique used in the original Australian earthquake loading code (AS 2121).
The same ground motion attenuation function was used in both techniques but for a direct comparison the aleatory variability was set to zero in the probabilistic case because the historical approach did not include this effect.
In the historical approach the variability in completeness of the recorded catalogue was not considered. It was simply assumed that all earthquakes producing accelerations greater than a given value would be recorded over the last 100 years.
The comparisons were made for minimum considered magnitudes of 4 and 5.
There was general agreement between the two approaches especially at shorter return periods (lower PGA amplitudes). At longer return periods (higher PGA amplitudes) where there were higher uncertainties, the results at some sites diverged.
This simple comparison of two approaches to the same problem of estimating earthquake hazard is shown to be of value in ensuring that the AUS5 model used by SRC is producing results that are consistent with the historically recorded data.
Gary Gibson and Vicki-Ann Dimas
Earthquake recurrence models are based on observed seismicity, geological data and geodetic motion. They are particularly difficult to define in regions of low seismicity where the average recurrence interval between moderate to large earthquakes greatly exceeds the duration of the known earthquake catalogue.
The earthquake process may be considered as ongoing long-term deformation due to plate movement in the region about the fault, resulting in stress build-up, and a significant number of small earthquakes through the deformed region. Larger earthquakes occur at irregular intervals, with ruptures on the larger faults that release elastic strain energy from the region. Most strain energy release is during the large fault rupture.
This gives a wider range in hazard estimates compared with extrapolation methods, increasing hazard in regions of active faulting and reducing hazard where long-term geological stability can be observed. As dams are usually in regions with recent uplift, this method will tend to increase hazard estimates.
Gavan Hunter, David Jeffery and Chris Kelly
Laanecoorie Reservoir, located in central Victoria, passed 3 significant floods in late 2010 to early 2011; the last flood being the highest on record since 1909. Significant cracking and deformation of this 100 year old puddle core earthfill embankment occurred. A series of longitudinal cracks up to 25 mm in width opened up in the crest over a length of 70 m and crest settlements were up to 70 mm; very large for a dam of this age. A significant difference at Laanecoorie compared to other similar dams is that it experiences high tail water levels during major flooding.
Investigations into the embankment following the January 2011 flood encountered several defects
including a decomposed tree root hole (large void up to 90 mm) that almost fully penetrated the raised section of puddle core, permeable gravel layers within the puddle core and transverse cracks up to 2 mm wide. The encountered defects and performance of the embankment many years after construction highlighted the deterioration that can occur with aging of these older embankments and the issues associated with poor past practices in tree management adjacent to dam embankments.
Dam safety upgrade works were undertaken in 2013 to address the identified piping and stability risks.
The works included construction of a filter buttress, replacement of a length of the raised puddle core and construction of a buried gabion wall on the left abutment to provide protection against scour should the secondary spillway fail or overtop.
GMW implemented a series of actions during the flood events in accordance with the Dam Safety
Emergency Plan (DSEP) to address cracking and deformation. Once aware of the dam safety risks, interim actions were implemented including increased frequency of monitoring, together with set up and measurement of crack pins, and temporary survey markers on the embankment.
JN Rossouw, AHM Görgens and PC Blersch
Shallow lakes or reservoirs generally exist in either of two stable states; a clear water state dominated by rooted water plants, or a turbid state dominated by free floating algae. A dramatic event can switch a shallow reservoir from one state to another. Voëlvlei Dam, a relatively shallow off-channel storage reservoir in the Berg River catchment, South Africa, switched from a stable, clear water system to a turbid, algal dominated system when it was severely drawn down during a drought in the mid-2000s.
It appears that there is tipping point beyond which a shallow reservoir can switch from one stable state to another and that there are buffers that maintain it in a specific state. Voëlvlei Dam is a good example of what such a switch might be (low water levels and high wind mixing) and what buffers (change to bottom-feeding fish species) may maintain it in the new state. It is only by understanding the hydrodynamic behaviour of a shallow reservoir that one can predict what these switches and buffers could be. Complex hydrodynamic modelling and comprehensive fish monitoring will facilitate more informed decision making and better management of reservoirs.
This paper describes the mechanisms that lead to the switch and how it can be prevented by developing an understanding of the hydrodynamic behaviour of shallow reservoirs through hydrodynamic water quality modelling.