The 48-hour delay that sealed Brisbane's fate

ENTRIES reveal that the senior engineers in charge knew by 7.10pm on January 9 that high releases of water from Brisbane's Wivenhoe Dam would be needed.

ENTRIES in the "flood event log" for the devastating Brisbane River flood reveal that the senior engineers in charge knew by 7.10pm on Sunday, January 9, that high releases of water from Wivenhoe Dam would be needed "in view of heavy rain over the last three hours".

The entries also show that senior engineers proposed more than doubling the releases, from 1400 cubic metres a second (cumecs) to between 3000 and 3500 cumecs that Sunday night to give the dam more storage capacity to manage the flood and intensifying rainfall.

But it took until Tuesday, January 11, when the dam was almost full for SEQWater, which employs the engineers who operate Wivenhoe Dam, to start releasing more than 3000 cumecs.

By Tuesday evening, with the dam at risk, the releases were dramatically ratcheted up to 7500 cumecs - flooding thousands of Brisbane homes and leaving a damage bill of billions of dollars.

Revelations in the official log entries by key personnel of SEQWater during the disaster raise serious questions about whether engineers maintained a status quo of low releases after being told by the Brisbane City Council late on Sunday that higher releases would flood hundreds of low-lying homes.

Evidence in the flood event log points to the council's input influencing the release strategy on Sunday, January 9, when there was time to manage the flood.

This will be seized on by insurers and flooded residents amid calls for the log to be closely scrutinised by a public inquiry.

Independent engineers and other experts examining the saga believe they can demonstrate to the royal commission-style inquiry that most of the flooding in Brisbane was caused by poor management early on and then massive sudden releases from the dam on Tuesday, January 11.

Senior independent engineer Michael O'Brien, who is compiling a submission for the inquiry headed by Supreme Court judge Cate Holmes, told The Australian last night: "It is clear from reading the flood event log that the flood operations centre knew as early as 7.15pm on Sunday, January 9, that much larger releases from Wivenhoe, of about 3000 cumecs, were required from as early as midnight and that this was important enough to notify the director of dam safety. But releases at this rate did not occur until 10am Tuesday, which by then was just too late. Even though the flood event log shows continuous discussions about the need for substantially increased release rates from this time, the discharge rates were increased too slowly."

The log entries reinforce the concerns of senior independent engineers and investigations by The Australian that much of the river flooding that destroyed thousands of homes occurred due to SEQWater's strategy to hold on to too much water in the dam.

Examination yesterday of the log entries and situation reports show that for most of January 10 the releases were kept between 1400 cumecs and 2000 cumecs - more than 30 per cent below the figure proposed by engineers the previous evening. By 6am the following morning, the dam was almost full, while the releases were still just 2750 cumecs. At 6pm that day, with SEQWater and Premier Anna Bligh gravely concerned that the dam was nearing its safe maximum capacity, the releases were increased to about 7500 cumecs and this resulted in most of the flooding.

Revelations in the log entries are not reflected in the main body of a newly released 1180-page report by SEQWater, which concludes the agency performed well in following the operating manual and the dam mitigated a worse flood. But the flood event log shows that at 7.15pm on January 9, an officer in the Flood Operations Centre "called SEQWater CEO Peter Borrows advising him that high rainfall is expected overnight and releases from Wivenhoe causing damaging flooding are likely to be necessary". A 7.15pm entry shows dam personnel called the director of dam safety to advise they were "now looking at much larger flows and will have to ramp up releases to around 3000 cubic metres per second by as early as midnight, which is likely to have flooding impacts on low-lying areas of Brisbane". Another entry five minutes later shows that "Engineer 2 called (Brisbane City Council) advising him of potential for high releases sooner than previously expected".

The entries show that after hearing from council that "3500 cumecs is the damaging flow level for Brisbane urban areas", the engineers in charge decided to keep releases from the dam at less than half the proposed rate, or just 1400 cumecs.

An entry in the earlier hours of January 10 documents an engineer's call to the dam operations manager "to discuss (council's) view on damaging flow. Engineer 3 confirmed that if flows were kept below 3500 the fuse plug would be triggered".

This 12.55am entry shows that at a relatively early stage, engineers knew a potentially disastrous outcome, a collapse of the auxiliary spillway or fuse-plug, could occur due to massive rainfall run-off unless releases immediately increased past 3500 cumecs.

Within 40 hours of this prediction, the rainfall and run-off had filled the dam to a level approaching the fuse-plug, resulting in the operators on the afternoon of January 11 ordering the release at 7500 cumecs.

SEQWater has a statutory requirement to operate Wivenhoe Dam independently and in accordance with an operating manual.

The official report justifies SEQWater's decision to not release more water sooner by pointing to Bureau of Meteorology forecasts, which underestimated rainfall intensity.

A spokesman for the Bureau of Meteorology referred questions to the floods inquiry, ignoring assertions from SEQWater that the forecaster was responsible for inadequate rainfall warnings in the lead-up to the January floods.