Melbourne Water (MW) has historically seen dam safety management as a civil discipline and has focussed on understanding and managing the civil assets at its dam sites. The recent addition of a mechanical engineering resource to the team responsible for the dam safety management has refocused attention on the mechanical and electrical (M&E) assets and provided a more holistic asset management approach to MWs large dams.
This paper discusses the process MW has developed over the past two years to improve their understanding and management of M&E assets. It centres on key process points for how MW has prioritised the development of M&E asset management programs on the basis of an autogenous ‘asset criticality’ rating system and has utilised ANCOLD comprehensive inspections to plan and implement new inspections and tests on dam M&E assets. The two case studies of Sugarloaf and Upper Yarra Reservoirs’ outlet works demonstrate the the benefits of the process to gain operational and technical knowledge of M&E assets, strategic importance to the water supply network, identifying risks therein and reallocate significant funding to address these risks as prioritised by asset criticality.
Now showing 1-12 of 59 2982:
The U.S. Army Corps of Engineers (USACE) is responsible for flood risk management across the United States. USACE has more than 710 dams and is responsible for more than 24,000 kilometres of levees. Since 2008, USACE projects have prevented more than AU$1.2 Trillion (in 2017 dollars) in damages from flooding. Although some of this came as a result of dozens of smaller floods, much of that protection came during three events within the last five years. From 2010 through 2017, the U.S. has had three major inland floods and two coastal events where federal flood protection exists: in 2010 on the Cumberland River, in 2011 on the Missouri, Ohio, White, and Mississippi Rivers, in 2015 on several rivers in Texas and Oklahoma, and in 2017 along the Gulf Coast of the U.S. and its territories in the Caribbean. For many of these locations, these events produced record rainfall and the flood of record. USACE operated many large facilities on these systems and those systems overall performed as expected. However, USACE also experienced some operational issues, did a substantial amount of flood fighting, had several incidents, and several failures. This paper will describe the flooding experienced in those events, the operations of the flood protection systems, the performance overall, and some of the lessons learned.
Trustpower is a New Zealand based hydro generator and retailer. It started off as a business that only owned a few schemes and then during a period of rapid expansion between 1998 and 2002 acquired the bulk of its current schemes. Now it owns and operates 25 hydro schemes across New Zealand ranging from 150kW to 80MW output.
This paper examines how Trustpower’s Dam Safety Management System (DSMS) has evolved over time, taking account of developments in the business environment, proposed regulatory changes, improvements in the NZSOLD guidelines and evolution in international dam safety practice.
The Kumara-Dillmans-Duffers Hydro Electric Power Scheme (HEPS) and in particular its Kapitea Reservoir (high Potential Impact Category) will be used as an example to highlight how the DSMS evolved over this period.
Lake Buffalo located on the Buffalo River near Myrtleford in Victoria was constructed in the 1960s as a cofferdam for the then proposed Big Buffalo dam. Consequently, the dam was designed for a short life (<10 years) and design features and criteria for a permanent dam were not implemented.
Critical features include a primary spillway with three vertical lift gates, two outlet conduits located
through the spillway piers, a single upstream valve on each outlet conduit for regulation and isolation, and a multi-part bulkhead which is installed in front of the valves for inspection and maintenance.
With the continued operation of the dam beyond 60 years, upgrades appropriate to a permanent dam have been implemented, including addressing deficiencies with spillway gate hoists lifting equipment and redundancy of the outlet conduit vales. This proved challenging, as the operation of spillway structures does not readily align with industry or Australian Standards. This paper will outline the issues encountered, their resolution and the lessons learnt during this upgrade work.
Physical modelling of dam structures remains a preferred method for validating and improving dam designs. Flow behaviour in the approach and over the crest of a dam can be accurately studied with traditional methods such as pressure transducers, piezometers and current meters due to the relatively smooth and steady flow conditions. However, characterising flows within a stilling basin is far more difficult due to the complex, aerated and highly turbulent flow conditions. Recent work on detailed measurement of hydraulic jumps using a line-scanning Lidar was adapted for measurement of stilling basin surface profiles in a 1:50 scale model of Somerset Dam, QLD. Lidar was shown to be an effective and efficient tool for providing assessment of the toe jump, boil and flow into the downstream channel.
There are a number of software packages that have been developed to conduct Probabilistic Seismic Hazard Assessments (PSHA’s). Each one has advantages and disadvantages. Two such programs are compared; the licenced subscription-based EZ-FRISK software package developed by Fugro USA Land, Inc. and the open-sourced OpenQuake-engine (OQ) software package by the Global Earthquake Model (GEM) Foundation. Both of these packages use the classical PSHA methodology as described by Cornell (1968) and modified by McGuire (1976). Each of these packages offers different advantages; OQ is freely distributed, code based and provides easy access to a number of tools. EZ-FRISK doesn’t rely on command-line tools and instead provides an easy user interface with quick access to plots to check results. EZ-FRISK is computationally faster than the OQ program.
A simple rectangular source model with four sites was used to investigate the degree of agreement between these two software packages. Results indicate that hazard estimates from the two packages agree to within 4% for the two closest sites. At long return periods for the two furthest sites, the difference is larger.