Queensland and NSW are recovering from record breaking floods and again many are questioning the state of flood mitigation in Australia. But while focus remains on the disaster in Queensland, Sydney’s Hawkesbury-Nepean system could be the next big flood risk.
The Hawkesbury-Nepean has a history of flooding; floods in 1799, 1800, 1801, 1806 and 1809 caused substantial agricultural losses in the early settlement. In 1810 Governor Lachlan Macquarie enacted what was probably the first disaster mitigation measure (pdf) in the country’s history: creating towns on the high ground and encouraging the colonists to move there. Macquarie didn’t know that the high ground on which he established Windsor, Pitt Town, Richmond, Wilberforce and Castlereagh was still prone to flooding in rare events.
The Hawkesbury-Nepean valley has been likened to a bathtub — able to capture huge volumes of water but only release them slowly. The most extreme flood possible could reach a height of 26 metres at Windsor, the valley becoming an inland sea and submerging entire houses under many metres of water. Damage from such a disaster would cost close to $12 billion, destroy more than 15,000 homes and affect nearly 93,000 people. With current infrastructure 22,000 people could be unable to evacuate in time and be trapped in the floodplain and could drown under the rising waters.
Whether the Hawkesbury-Nepean could be evacuated fully is unclear, but the track record is unedifying. In 2001 authorities called for the evacuation of Grafton due to predictions that floods were going to burst the town’s levee. It’s estimated that only 13 per cent of the population left their homes. In 2005 evacuations in Lismore and Byron Shire had response rates (pdf) of 40 per cent and 19 per cent respectively.
Failure to evacuate during a flood in the Hawkesbury-Nepean could trap thousands of people on "shrinking islands" that will ultimately be inundated, drowning those that remain. Heroic rescue efforts will be hampered by swift flowing waters and poor weather — making it unlikely all those trapped will be reached. Hundreds or even thousands might die.
The scale of the potential catastrophe was only understood in the 1990s when research on extreme rainfall revealed the possibility of catastrophic failure of Warragamba Dam. Raising the dam was investigated as a solution to the possibility of an extreme flood, and also as a possible safeguard against the dam breaking. The government of the day decided to focus on the protection of Warragamba Dam by constructing a new spillway. Flood risk was to be managed by improving evacuation infrastructure.
Released in October last year, Infrastructure NSW’s State Infrastructure Strategy has placed flood mitigation in Western Sydney back on the agenda.
In a report prepared for the strategy, consultants Molino Stewart have concluded that raising Warragamba Dam by 23 metres would substantially reduce the flood risk in the valley and have a very high cost-benefit ratio.
However, these benefits only accrue for existing development conditions. Were planning restrictions to be relaxed and development intensified, the expected reduction in damages could evaporate. Governments have difficulty resisting development pressures after mitigation infrastructure like dams or levees are built. In the United States inappropriate development behind levees is rife and some consider it to be the driving factor behind increasing flood losses.
In an independent review (pdf) of Brisbane City Council’s 2011 flood response, the review board noted that the council had set its flood planning level at the height of the 1974 flood — after also deducting the capacity of Wivenhoe Dam.
"[I]t is considered that the January 2011 flood event, as actually experienced, was larger than a flood similar to that of 1974 after mitigation by Wivenhoe Dam," the report commented. Despite peaking a metre lower (pdf) 9300 more homes were flood affected in 2011.
The fringes of the Hawkesbury-Nepean floodplain have been targeted for development by previous NSW governments. This could increase development pressure further into the floodplain.
Many councils have a current planning level for residential development equivalent to 17.2 metres at Windsor; hardly an improvement since the 1970s, when levels were at least 16 metres. Raising the dam would reduce flooding so that an equivalent level would be 4.4 metres lower. If adopted as a new planning level it would allow development at levels that are below all but the oldest dwellings in the valley. It’s not clear how much additional development would be possible with this change, but it could be substantial enough to eliminate the benefit of raising the dam.
The risk of an extreme flood that would require the evacuation of the whole valley remains even with a higher Warragamba Dam.
The NSW State Emergency Service has prepared a comprehensive evacuation plan (pdf) for the Hawkesbury-Nepean. It anticipates that in an extreme flood there may be as little as 24 hours to start and complete an evacuation before roads are cut by floodwaters. Raising Warragamba would delay the arrival of floodwaters long enough to increase time available for evacuation so that even with present infrastructure it would be possible to get everyone out. Upgrading road infrastructure is another possible option.
But people often put too much trust in mitigation works; belief that Wivenhoe Dam would protect Brisbane and Ipswich was termed the "Wivenhoe Effect" during the Queensland Floods Inquiry. It may have contributed to the losses and heartache from that event. A "Warragamba Effect" could see thousands of residents not heeding evacuation warnings in an extreme flood. Combine that with increased development and the number of people failing to evacuate could lead to a potentially higher death toll in a catastrophic flood than is presently possible.
Future development pressures aren’t the only thing that could undo the benefits from raising Warragamba. A new study has found a strong link between climate change and extreme rainfall. The implications of this research on the Hawkesbury-Nepean are unclear but global warming could further erode the benefits of raising Warragamba Dam.
In its response to Infrastructure NSW the NSW Government has committed to reviewing (pdf) "the major flood mitigation options available in the Hawkesbury Nepean Valley, including the options of raising the Warragamba Dam wall and road upgrades". The timeline for this review is unknown, as is how it will address the risks from a catastrophic flood.