The 2005 annual US Association of State Dam Safety Officials (ASDSO) Dam Safety conference was scheduled to take place on 25 to 29 September in New Orleans, Louisiana. The tragedy of Hurricane Katrina in August however forced ASDSO to rearrange the event, changing the time and venue. The damage caused by the hurricane to the city and its infrastructure meant there was no way the original venue could be used.

ASDSO, which ran regular updates on its website, moved the conference east to Walt Disney World in Orlando, Florida, from 23 to 27 September. Most of the delegates and speakers were still able to attend, and the yearly event commenced as normal, but perhaps under more sombre circumstances than usual.

Hurricane katrina

As well as rescheduling and moving the conference, Hurricane Katrina’s influence carried over into the event itself. The nature of the disaster meant that it would be impossible to ignore it in a meeting about dam safety – the collapse of New Orleans’ levee system and the damage from the below sea level city’s watery borders had been analysed at length worldwide. In a charitable gesture, ASDSO set up a ‘silent auction’ (bids were written down on paper) for items such as signed sports memorabilia and DVDs, supplied from representatives from each of the 50 US states – the money going towards the Hurricane Katrina Relief Fund.

Dam Safety 2005’s opening proceedings included a Hurricane Katrina update by US Army Corps of Engineers (USACE) representative Eric Halpin as part of a general session on levees. Halpin gave a perspective on Katrina at that time, emphasising the tragic loss of life that occurred in the affected areas, and the help that groups like USACE, the US Bureau of Reclamation (USBR) and the Federal Emergency Management Agency (FEMA) gave.

Halpin presented a perspective on New Orleans and the hurricane. He pointed out that the city is 3m below sea level – a fact that has always made it susceptible to flooding from the surrounding lake Pontchartrain to the north (which at 64km wide is more than twice the city’s size) and the Mississippi river to the south. The storm was particularly bad – a class IV hurricane, and the levee defence system was built to withstand up to a class III storm only. It was, as Halpin put it, ‘a very large impact on a very large system’. There was damage not only to the 120km flood protection system itself, but also the interior drainage system and the pumps that were designed to aid in a flooding situation. The storm surge pushed boats ashore; in some cases there was overtopping, with barges actually pushed on top of the levees. It was difficult to get in by land, so helicopters were used to drop sandbags in. Despite this, New Orleans actually became dry ahead of the predicted schedule.

Currently, USACE has set up the Interagency Performance Evaluation Task Force (IPET) to undertake comprehensive analysis to determine exactly what happened to New Orleans and its flood protection system during the hurricane. IPET is comprised of some of the nation’s leading engineers and scientists from government (federal, state and local agencies), academia and private industry.

Furthermore, USACE announced that the Assistant Secretary of the Army for Civil Works has decided that rebuilding the New Orleans levees will be at full federal expense; the coordinated restoration of the hurricane, flood and storm damage reduction projects to pre-storm conditions will be provided. Data about the New Orleans flood protection system, with an emphasis on an engineering perspective, has also been made available on the USACE website (

Satellite dam monitoring

One interesting paper from the conference was ‘Use of GOES satellite transmissions in dam monitoring’ by Thomas N Keefer, Vice President of Special Projects, Sutron. In the paper, Keefer explained that all USACE dams are required to have a level of instrumentation that enables proper monitoring and evaluation of the structures.

Focusing on 31 flood control dams operated by the New England district of USACE, Keefer explained that traditionally, information on piezometer levels and other geographical parameters has been collected by hand or by ground-based radio or hard-wire telemetry systems. However, Sutron worked in association with USACE to develop alternative telemetry networks combining line-of-site radio, hard-wire and Geostationary Operational Environmental Satellite (GOES) messaging for the dams in the US states of Connecticut, Massachusetts, New Hampshire and Vermont.

GEOS satellites are the worldwide network for sending information for the real time monitoring of weather. Utilising them to monitor piezometer networks, Keefer insisted, creates a much more comprehensive dam monitoring system, designed to improve the early warning of potentially critical embankment performance parameters. Typically, the dams transmit the pool elevation and outlet discharge, but the GOES telemetry devices, called data collection platforms (DCP), can automatically take readings from any of the common hydrologic instruments used for pool elevation, stream stage, and weather.

The geotechnical instrumentation at the dams consists of piezometers, seismic strong motion instruments, crest survey monuments, tilt plates, inclometers and other instruments that are deemed necessary to adequately monitor embankment performance. Automated instrumentation systems are also currently being installed at each dam to collect and transmit data from the piezometers, seepage measurement devices, pool level sensors and strong motion instruments direct to the New England office.

Sutron was responsible for developing custom made software for the loggers at each dam site’s master data collection station’s programmable data logger. The data collected at the sites is sent via either modem and/or GOES transmissions and is streamed into a database. Dam safety information, in the form of graphics, graphs, tables or actual visual images, can then be posted onto the internet.

The paper lists some problems that were encountered along the way – such as bringing signals from a number of individual piezometers to the central DCP location, and interfacing all the sensors and vibrating wires with a DCP – but these were overcome, and Keefer concluded by saying that the GEOS system has reached an impressive level. At some of the dams it is possible to view piezometer levels each hour, and this kind of efficiency should help the New England office meet its mandate to have a level of instrumentation that enables proper monitoring and evaluation of the dams under all operating conditions.

Potential failures modes analysis

‘Don’t judge the book by its cover: experiences with potential failures modes analysis’ was a paper written by Greg Hammer, a dam safety engineer of the Greeley, Colorado Division of Water Resources, Dam Safety Branch. It concerns efforts by Colorado Dam Safety Branch personnel to develop a standardised methodology for more efficient evaluations of potential and current dam safety problems. A pilot programme was implemented in the summer of 2001 to utilise the potential failures modes analysis (PFMA) evaluation concept using its failure modes and consequence analysis (FMCA) technique to identify conditions that could lead to a dam being compromised or failing. This was followed shortly after by another pilot programme, the Failure Assessment Index (FAI), an internal methodology for storing and analysing the data that is designed in conjunction with the USBR’s recently-developed Risk Profile tool to prioritise which dams to select for FMCA evaluation.

Hammer said that the significance of this project was how it changed the way that he and other dam safety personnel began to see the issue of dam safety – primarily that looks can be deceiving. Despite efforts to make a dam safer, actions taken can leave it still attaining the same level of inherent structural integrity – and in some cases, efforts in caring for’ a dam can actually be detrimental rather than beneficial; a problem that tends to occur if standard policies only are followed, rather than each dam being seen as a unique case.

Under the guidance of the USBR’s Larry Von Thun, the pilot scheme used FMCA over the summer to study eight dams. On one dam, labelled ‘S’, a potential problem was identified in that the large canal around the structure might collect storm inflow from a non-tributary basin, leading to overtopping. This was considered a high priority, as not only did the canal begin several kilometres away from the reservoir – so it could potentially collect storm inflow from many basins – but the dam was also close to a heavily populated area, meaning that an overflow could have potentially devastating effects.

Another dam, labelled ‘R’, had a problem that had already been observed – seepage. Several locations had experienced the problem, such as the downstream slope of the dam. The reason for the seepage was largely put down to a less impervious layer of material creating a perched condition in the embankment, and continued monitoring took place to assess the problem at this and other locations. However, it was during the course of applying FMCA monitoring modes that a new concern was recognised – fractures and joints of the abutment bedrock had been exposed to erosion. This situation had generally been overlooked due to attention being focused on the other problems, but during the course of the FMCA discussion it was decided that the potential for distress was higher in this new area, even though no previous signs to suggest this had been recorded.

In the paper, Hammer suggests that PFMA has offered the Colorado Dam Safety Branch a better understanding of the process of dam failures. Having used these methods, he said, they now understand the need to focus attention toward a specific understanding of what might be present below the ground surface, and not simply considering the visual observations; continuing the metaphor of the paper’s title, he said that it is crucial to see past the book’s cover and read what is inside.

Pearl Harbour and 9/11

The second day of Dam Safety ’05 began with the the intriguingly titled paper ‘Complacency can’t win: The systems were blinking red – what Pearl Harbour and 9/11 teach us about dam safety’, from Todd E Martin of AMEC.

Martin began by pointing out that 79,000 dams in the US did not fall down in 2004. However, we do not hear on the news about all the projects that held firm; instead only the catastrophes and failures are reported. ‘The dam safety industry has always been defined by its failures,’ was how he put it. He also emphasised that this is no longer acceptable, and parties involved in dam safety cannot afford to be merely reactive.

Indeed, in his paper he put forward the opinion that failures of dams can inflict greater damage to life and property than any other type of civil engineering work, making the responsibility of those involved a huge one. He pointed out that not only is physical damage caused, but drinking and irrigation water supplies can be affected and if the local area has been relying on the structure’s hydroelectricity, than its entire power generation capabilities will be brought to a standstill.

His paper was about Pearl Harbour and 9/11, but in the presentation at the conference he used the topical Katrina tragedy as a comparison too. In the paper, he insists that surveillance and emergency response planning are as integral to dam safety as they are to military intelligence and national defence. With this in mind, he sees the strike on the Hawaiian military base on 7 December 1941 during World War II and the New York Twin Towers air attack on 11 September 2000 as providing invaluable and sobering case studies to be related to dam safety. He says that both events were results of failures in intelligence management and contingency response planning, and they were both studied at length in the US to try to learn from any mistakes. But were lessons really learnt – and will they be from Katrina?

The paper presents water as a great opponent: ‘relentless, infinitely patient, capable of concealing its attack, and capable of finding and exploiting any weakness in a dam’s defenses’. Acknowledgement of this potential for destruction is evident in the amount of thought and money that goes into the design, materials, construction, monitoring and upkeep of any modern dam, and also by the very existence of events like the annual ASDSO conference . Martin points out that this opponent, at once a necessary friend and evil, can be restrained but not deterred, and it is regrettable but understandable that the extensive periods of calmness can breed a complacency that can be shattered when things, literally, come tumbling down.

And complacency is the major main mistake history reminds us about at Pearl Harbour and 9/11. The former saw Japanese carrier-based planes take the US Pacific Fleet completely by surprise, the air strike succeeding despite a wealth of intelligence collected and decoded by US military intelligence that suggested a Japanese attack on the Pacific was highly likely. In the case of 9/11, Martin says that history has not yet settled on a verdict as to whether this was a preventable attack, but what it does represent is the US being struck with a devastating ‘sneak attack’ that the country had promised its people would never be repeated post-Pearl Harbour.

Martin says that in much the same way, the worldwide hydro power community must learn to not become complacent with preventing water ‘attacking’; it must always be not just aware, but alert, innovative, quick-thinking and capable.


Anyone involved in the construction, maintenance or running of a hydroelectric project knows that safety has to be the number one priority. Safety measures generally go unnoticed if they working; if a dam holds back its water it is perceived to be doing its job properly. However, as the New Orleans event demonstrated, a structure may look sufficiently sturdy but can still fail. It often takes a crisis to be a catalyst for change, and complacency can set in until there is a real human tragedy to jolt the senses.