Susantha Mediwaka, Nihal Vitharana, Badra Kamaladasa
Nalanda dam is the oldest concrete gravity dam on the Island built in the 1950s by the Ceylon Department of Irrigation. The dam was built in 9 monoliths having a dam crest length of approximately 125m and a maximum height of about 36m. The spillway consists of: (1) a low-level uncontrolled ogee-crested horse-shoe section with a crest length of 46m, and (b) a high-level broad crested weir with a crest length of 43m.
It was designed and constructed according to the then standard practices adopted throughout the world. Over the years, Nalanda dam has been showing signs of deterioration which is suspected to be Alkali-Aggregate Reaction (AAR). The dam was also shown to be deficient with respect to the stability levels required by modern standards. Under a program of dam safety improvement of the dams throughout Sri Lanka, it was decided to stabilise Nalanda dam as the first step in addressing a series of issues affecting the dam.
This paper presents the construction history, current issues, design assumptions and salient construction features in the upgrading of the dam to modern dam safety requirements.
Keywords: Concrete dams, dams Sri Lanka, concrete buttressing, upgrade, horse-shoe spillway
J.H. Green; C. Beesley; C. The and S. Podger
Rare design rainfalls for probabilities less frequent than 1% Annual Exceedance Probability (AEP) are an essential part of spillway adequacy assessment as they enable more accurate definition of the design rainfall and flood frequency curves between the 1% AEP and Probable Maximum events.
Estimates for rare design rainfalls were previously derived using the CRC-FORGE method which was developed in the 1990s. However, as the method was applied on a state-by-state basis, there are variations in the approach adopted for each region. Differences in the cut-off period for data, the amount of quality controlling of the data undertaken, the base used for the 2% AEP estimates, gridding settings and smoothing processes have created inconsistencies which are particularly apparent in overlapping state border areas.
The Bureau of Meteorology has derived new rare design rainfalls for the whole of Australia using the extensive, quality-controlled rainfall database established for the new Intensity-Frequency-Duration (IFD) design rainfalls. These data have been analysed using a regional LH-moments approach which is more consistent with the method used to derive the new IFDs and which overcomes the limitations of the spatial dependence model in the CRC-FORGE method. In particular, the selection and verification of homogenous regions and the identification of the most appropriate regional probability distribution to adopt relied heavily on the outcomes of the testing of methods undertaken for the new IFDs. However, to focus the analysis on the rarer rainfall events, only the largest events have been used to define the LH-moments.
Keywords: Rare design rainfalls; Intensity-Frequency-Duration (IFD); Annual Exceedance Probability
Gavan Hunter, Andrew Pattle and Mark Foster
A piping incident occurred during first filling of Rowallan Dam, Tasmania in 1968. The incident occurred at the interface of the embankment with the spillway wall, a 15 m high near vertical wall, where the contact earthfill eroded into the single stage downstream filter. Repairs were undertaken in 1968/1969 and the reservoir has operated largely without incident since.
A risk assessment in 2009 identified that piping through the embankment at the spillway wall interface remained a significant dam safety risk. Investigations in 2010 encountered cracking within the earthfill core at the spillway wall interface.
Dam safety upgrade works were undertaken in 2014/15 to address the piping failure mode at the spillway walls and also within the upper portion of the embankment. The works required excavation down to a rock foundation at depths up to 18 m adjacent to the spillway walls and this excavation provided an unusual opportunity to closely examine active piping features that had been preserved when interim repairs in 1968/69 had arrested the progression of piping. The repair comprised reconstruction of a significant portion of the embankment at the spillway and the reconstruction of the upper 7 m of the crest, which included dual filters downstream of the earthfill core.
The findings from the forensic investigations of the deep excavations adjacent to the right spillway wall are described in this paper along with a summary of finding from the 1968/69 repair works and a discussion of the piping mechanism at the spillway wall. The paper also covers the design and construction of the repair work. The focus of this paper is on advancements in our understanding of piping risk arising from the Rowallan Dam work.
In conclusion, (i) the upgrade works successfully reduced the dam safety risk of Rowallan Dam; (ii) the findings support the methodologies of the piping toolbox; (iii) the case study provides insight into filtering and crack filling mechanisms that have a broader implication for estimating the risks of internal erosion within existing dams; and (iv) the findings support the assessment of the low residual risks for piping through the embankment away from the upgrade work areas (crest reconstruction and spillway walls).
Keywords: Earth and rockfill embankment, piping incident, piping mechanism, dam safety upgrade.
Malcolm Barker, Toby Loxton
The Gladstone Area Water Board (GAWB) owns and operates Awoonga Dam, which is a concrete-faced rock fill embankment with a fixed crest concrete spillway on the left bank impounding a storage volume of 770,000 ML.
The current arrangement can accommodate the Probable Maximum Flood, allowing for flow over Saddles 3, 4 and 6 on the left abutment. A comprehensive study was carried out to evaluate the erosion potential downstream from Saddles 3 and 6 as well as other spillway options adjacent to the existing dam. One option was a radical approach including the removal of the Saddle Dam 3 and provision of downstream erosion protection works. This reduced the PAR and improved the overall dam flood capacity; however concerns were expressed about the environmental impact of possible erosion downstream from Saddle 3 for relatively frequent events.
A risk assessment showed that the erosion protection works downstream from the Saddle 3 or 6 were not cost effective and the preferred option for the upgrade was the closure of the Saddle Dam 3 with an auxiliary spillway created in Saddle 6,
This paper summarises the methods used and the outcomes from this study.
2011 – Awoonga Dam Acceptable Flood Capacity design – the anguish of erosion risk and implications for design
Petros Armenis, Malcolm Barker, Peter Christensen, Graham Harrington
The Canterbury Earthquake Sequence in September 2010 and February 2011 caused large areas of land to change by differing amounts throughout Christchurch, New Zealand. Land levels fell by more than 300 mm in some areas. This increased flood risk in the tidal reaches of the Avon River. Urgent repairs were completed with the objective to restore the tidal river defences to a crest level equivalent to a 1% AEP tide level. This work needed to be completed prior to impeding spring tides.
The levees will be required for up to 20 years and then probably be rebuilt on a new alignment. To better understand the risks associated with the ongoing reliance of the levees for flood protection in the interim, a risk assessment was undertaken using conventional Australian National Committee on Large Dams (ANCOLD) practices and levee design procedures. Careful consideration was made to the performance of the existing levees under seismic, flood and tidal loading from which the societal and individual risk profiles were derived. The work included the following:
This paper will present the levee design and the process applied for the analysis of the levee and the upgrade options selection