Peter Hill and Rory Nathan
The ANCOLD Acceptable Flood Capacity (AFC) guidelines were published in 2000 and provide guidance on the selection of design flood capacities for dams and specifically a deterministic fallback provision for spillway capacities. Since the guideline was published, there has been a continual evolution in dam safety management practices and related guidelines, including the 2003 ANCOLD guidelines on risk assessment and the current revision of Australian Rainfall and Runoff by Engineers Australia. This paper describes the scope of the current AFC guidelines and perceived opportunities for refinement. A survey of users was used to test and identify issues and gauge the need for the guideline to be updated. A number of topics were identified that would benefit from clarification or further guidance. These topics include consistency with other ANCOLD guidelines, clarity on the selection of the AFC, definition of the dam crest flood, freeboard and application to gated structures.
Neil Jacka, Christopher Dann, Jeremy Eldridge
The Tekapo Canal Remediation Works were undertaken to extend the life of the canal and enhance its seismic and environmental resilience. The deterioration of the canal lining in specific reaches has been the consequence of internal erosion of the lining under operating conditions.
The remedial works comprised installation of a supplementary geomembrane liner over selected sections of the canal, reconstruction of a culvert where the embankment had suffered piping, installation of filters in the Maryburn Fill, strengthening of the bridges across the canal and replacement of irrigation off-takes.
This paper presents a summary of key issues resolved during the design of the remediation works, in particularly the design of the geomembrane ballast system, the cofferdams and the management of side slope stability during drawdown for the works. A number of construction trials were carried out to confirm design assumptions and test construction techniques. The trials were a significant factor in the successful completion of the first season of work ahead of programme.
Keywords: Canal, Lining, Geomembrane, Cofferdam, Design, Seismic resilience
Tim Gillon and Grant Murray
Chelsea Estate is located on the edge of the Waitemata Harbour, and is only ten minutes drive from Auckland central business district. Within Chelsea Estate are four ‘low’ potential impact classification (PIC) dams, which cascade along Duck Creek. Three of the dams are over 100 years old and all dams were built from 1884 to 1917. The dams and the reservoirs have served, and continue to serve, several purposes including stormwater retention, recreational use and water supply for the adjacent sugar factory. In 2008 Auckland Council (AC) purchased the Chelsea Estate from the New Zealand Sugar Company (NZSC) and in 2009 the Estate was registered in the New Zealand Historic Places Trust (NZHPT). This paper discusses the history and functionality of the multi-function Chelsea Estate dams, the development of the site and how it impacts our understanding of the dams today.
Keywords: Chelsea Estate, multi-function dams, heritage dams.
Kinchant Dam is a zoned earth and rockfill embankment situated on the north branch of Sandy Creek, approximately 30 km southwest of Mackay in central Queensland. Kinchant Dam was constructed in stages. The ‘Initial Development Stage’ which consisted of an embankment length of approximately 3.3 km and full supply level (FSL) of EL 49.21 m AHD was completed in 1977. Further development completed in 1986 (Stage I) increased the FSL to EL 57.21 m AHD with an embankment length of 5.5 km and a maximum embankment height of 22.3 m. The dam has a storage capacity of 62,800 Ml and a 60 m wide emergency spillway with a fixed crest level of EL 58.21 m AHD, one metre higher than the FSL.
A series of investigations have been carried out since its construction as a consequence of both regulatory safety reviews and observed excessive pore pressures within the foundation that have led to wet patches developing at the toe of the dam. In one area at the toe, pore pressures were such that artesian conditions developed. This paper outlines the history of various stages of construction of the dam, the foundation investigations since construction and the safety review and comprehensive risk assessment process that lead to the upgrade design and construction of remedial works. The remedial works include the extension of the downstream filter material adjacent to the clay core and the provision of additional pressure relief wells at the downstream toe of the dam.
Upstream construction methodology has been used to raise tailings dams in Western Australia (WA) for more than three decades, and the tailings storage facilities (TSFs) built in this manner have performed satisfactorily so far. The maximum design earthquake (MDE) for most of the existing, upstream-raised TSFs in WA was that corresponding to a 1-in-1,000 year annual exceedance probability (1:1,000 AEP). However, the recommended MDE loading for the High/Extreme Failure Consequence Category in the 2012 ANCOLD Guidelines on Tailings Dams is that of a 1:10,000 AEP. This more stringent seismic design criterion may restrict the use of upstream TSF construction in some areas of WA and Australia in general.
To evaluate the viability of upstream construction for a new or existing TSF, the effects of the earthquake design ground motion (EDGM) on the liquefaction and deformation response of the structure must be understood. The results of such analyses are an essential component in determining whether upstream raising will be feasible, or whether more robust but much more costly centreline or downstream construction methods are required.
A parametric study was completed to investigate the liquefaction and deformation behaviour of a typical, upstream-raised tailings dam under different earthquake design ground motions with different response spectra. The study utilized two-dimensional finite difference code FLAC2D effective stress dynamic analysis, in which the UBCSAND constitutive soil model was incorporated. Twenty-eight earthquake ground motions (matched and unmatched to the target response spectrum) were used in the study and the liquefaction response of the tailings dam model under those ground motions was analysed.
The results of the study demonstrate the importance of appropriate ground motion and response spectrum selection in assessing the seismic performance of an upstream-raised TSF. Liquefaction response was shown to vary with different response spectra, even though the corresponding EDGMs had similar peak ground acceleration (PGA) values. The importance of earthquake frequency content and duration, which in turn are affected by earthquake magnitude, distance and ground motion response, is emphasized. Scaling and matching the earthquake input motion to the uniform hazard response spectrum (UHRS) may result in overly-conservative design. Thus, selection of the most representative EDGM is essential to evaluating expected seismic performance for an upstream-raised TSF, and scaling or matching the earthquake input motions must be done cautiously.
Earthquake ground motions were developed for the Tekapo Canal Remediation Project, including both Canal and Bridge sites. This work involved the specifications of the parameters of active faults and seismic source zones, the development of an aftershock sequence, and the review and selection of suitable ground motion prediction equations. The seismic hazard at the project sites is dominated by earthquakes occurring on the Irishman Creek fault. The characteristics of an inferred active (unnamed) fault shown crossing the Tekapo Canal near Forks Stream and the hazard it poses to the canal were also assessed, and it was concluded that there was no need to further investigate it as part of the canal upgrade project. A probabilistic seismic hazard analysis was used to develop response spectra for mainshock events for the various return periods relevant to components of the canal system having different PIC categories. A deterministic seismic hazard analysis methodology was used to estimate the aftershock spectra. Depending on the PIC category, time histories were developed to represent the response spectrum for both mainshock and aftershock events at some canal sites.