Operator of dam 'invented' rain data

EXTREME rainfall so rare it happens on average once every 2000 years has been "invented" by the government operator of a major Queensland dam

EXTREME rainfall so rare it happens on average once every 2000 years has been "invented" by the government operator of a major Queensland dam as part of its explanation for releasing huge volumes of water that caused most of Brisbane's January flood.

The claim by SEQWater in its official report that a "one-in-2000-year" rainfall event occurred over the Wivenhoe Dam at a critical stage on January 11 has been widely reported in the media and cited by senior public servants to justify the near loss of control of the dam at the time.

But no such rainfall event was measured by any rainfall gauges. Instead, the claim was manufactured by SEQWater after it modelled the rapid rise of levels in the dam, repositioned rainfall data to an area immediately upstream of the dam, and then doubled it.

After extrapolating in this unusual way to achieve an extreme number, the SEQWater report states: "Rainfall of this intensity and duration over the Wivenhoe Dam lake area at such a critical stage of a flood event was unprecedented.

"The resulting run-off could not be contained without transition to (an operating strategy that led to the operator opening the dam's gate for huge releases)."

Senior independent engineer Michael O'Brien, who has spent the past nine weeks analysing the performance of the dam and SEQWater, said that while the rainfall was heavy, he did not believe it was extreme and he doubted it was ever close to the range claimed by the operator.

Mr O'Brien, who has mounted a strong case that the devastating floods in and near Brisbane would have been almost completely avoided with better management of the dam, said the one-in-2000-year event was an "invention" that could not be taken seriously.

He said: "To get the inflow rate, SEQWater had to manufacture an unmetered rainfall event over the dam, which was twice the size of any of the metered rainfall events, and this becomes a rainfall event with a one in 2000."

The technical report by SEQWater shows it relied on a manual gauge of dam levels, not the actual rainfall in gauges, to extrapolate data to claim the occurrence of a one-in-2000-year event.

However, in doing this, SEQWater disregarded the data from a nearby electronic gauge, which showed dam levels lower than those in the manual gauge.

Mr O'Brien said SEQWater's methodology in adopting the data from the manual gauge, ignoring the data from the electronic gauge, and then having to "scale this rainfall up by a factor of two to match the rapid lake level rises" would become a "major technical argument".

"When they calculated the dam inflow rates that would be necessary for the manual gauge board to be correct, they had very high inflow rates -- much higher than any other time during the whole event," he said.

SEQWater has commended the management of Wivenhoe Dam during the flood; however, the Floods Commission of Inquiry headed by Supreme Court judge Cate Holmes has received several reports from experts, hydrologists and engineers that raise serious questions.

A panel of hydrologists and engineers has categorised the Brisbane River flood as a "dam-release flood", meaning it was largely the result of massive releases. Official SEQWater data has highlighted concerns the operator held too much water for too long in the dam over a few critical days before the flood, then released extremely large volumes.