Why was the Wivenboe too full to save Brisbane?

More evidence that the Wivenhoe dam may have been too full to properly protect Brisbane during a La Nina:

MORE serious questions about Wivenhoe Dam were raised yesterday by a senior engineer who claims the Brisbane River flooding was avoidable.

The engineer, Michael O'Brien, said the official data on water flows shows last week's Brisbane River flooding would have been largely avoided if the dam operators had raised their releases of water on the weekend before last Monday's deluge.

Mr O'Brien, and others not directly involved with the dam's operation, told The Australian that the river flood and the devastation of thousands of homes was inevitable after a decision to release relatively low volumes of water on Friday, January 7, and over the ensuing weekend.

The data shows that the dam went from a little over 100 per cent of its capacity on Friday, January 7, meaning it still had capacity for more than one million megalitres in its flood compartment, to about 150 per cent by the morning of Monday, January 10 - before the deluge hit.

Over that weekend and prior to the extreme rainfall event that would flood Toowoomba and the Lockyer Valley on Monday, the dam's operators released a total of about 200,000 megalitres.

Scrutiny of official water-release and dam volume data shows the flood would have been moderate at worst in Brisbane had there been larger releases in the days before the deluge. But it became extreme due to the sudden surge from the release of about 645,000 megalitres from Wivenhoe last Tuesday, which was about 30 per cent of the total capacity of a dam built to protect the city and surrounding suburbs.

This sudden release came because the in-flow of water from the dam's catchment meant its reservoir had risen to alarming levels of about 190 per cent and were closing in on the level that triggers an uncontrolled discharge....But a more conservative approach over that crucial weekend would not have made any difference to the flooding that occurred in the Lockyer Valley and the city of Ipswich, both of which are in separate areas of the Brisbane River catchment.

SEQWater Grid chief executive Barry Dennien insisted last night that although the January 8-9 releases were relatively low compared with what occurred in the days afterwards, this was in accordance with the operating manual to mitigate flooding. He said that nobody had foreseen the extreme rainfall that ensued.

O'Brien's report concludes:

Hindsight is a wonderful thing; but there are really two decisions that were taken which would seem to have little justification at the time they were made. These are: -

o Not increasing releases from Wivenhoe between the 4th and the 10th when the dam levels were rising and there was no downstream flooding, and

o The decision to substantially increase the discharge rate from Wivenhoe to a peak of 645,000ML/d on the afternoon/night Tue 11th.

It is difficult to understand the justification for this increase in the discharge rate, especially as the decision was substantially reversed within about 12 hrs. In addition even if SEQWater had released at the low rate of 215,000ML/d (which is the rate it was reduced to at 07:30 Wed) rather than a maximum rate of 645,000ML/d for the whole time from 14:57 on Tues it would have used only 306,000ML of the available free capacity in both dams of apparently 859,000ML.

Read on for O'Brien's excellent report, giving discharge rates and times:

What went on in Brisbane?


Was this a natural disaster or a manmade disaster?


First a bit of a disclaimer – I am new to some of this river data and may not have interpreted it all


correctly.


First a bit of background: -

• Somerset Dam has a Water Storage capacity of 379,849ML with an additional 524,000ML for


flood storage. The dam is rated as 100% full when all the water storage capacity is full.


Therefore when both the water storage capacity and flood storage are fully utilised, the dam


would be at 238%.


• Wivenhoe has a Water Storage capacity of 1,165,238ML with an additional 1,450,000ML for


flood storage. Similarly when both the water storage capacity and flood storage are fully


utilised, the dam would be at 224%.


• Somerset is upstream of Wivenhoe and flood flows from Somerset are discharged into


Wivenhoe.


• There is a river height monitor at Wivenhoe Dam: -Station Number: 540177 Name: Brisbane


R at Wivenhoe Dam Hw # Owner: SEQWCO:143822

SEQWater operate a web site which gives levels in all dams, including Somerset and Wivenhoe;


but interestingly, levels for Wivenhoe and Somerset were not updated between 08:00 11th


January and 09:00 13th January during the peak of the emergency. Historical data now


shown for this period on the SEQWater web site does not show any peak during this period and so


are obviously incorrect.


I have therefore used the Wivenhoe Dam river height monitor as a proxy for the dam level during


this period. There are some interesting disconnects between the river level data and the reported


water levels in Wivenhoe: -


Information from media reports had Wivenhoe at a peak of 191% overnight for the night of


11th/12th; but generally at 190% through most of the emergency period.


The river height shows a height of 73.77m at the time SEQWater were reporting a Wivenhoe


capacity of 175.9%. The river experienced a minor peak of 74.51m commencing at 14:57m


Tue 11th falling substantially (to 74.19m around 17:45 Tue 11th) until peaking again at


74.85m between 18:00 and 19:23 on Wed 12th (compared to a reported peak in Wivenhoe


during the night of 11th/12th). The river level fell gradually, and has continued to fall, from


this peak.


The sequence of events for the current flooding event seems to be: -


• 06:30 Tue 04th, Wivenhoe first went above 100% (i.e. using the flood storage capacity).


• Wivenhoe was at 106.3% at 06:00 on Fri 7th; but there are no reported heights for Sat 8th or


Sun 9th.


• Wivenhoe continued to rise, and at 09:00 on Mon 10th it was at 148.4% and it was reported


that "managers scrambled to increase the release from 116,000ML to 170,000ML per day.


• At 0800/0900 Tue 11th Wivenhoe was at 175.9%, Somerset at 160.8%; total available


capacity for flood storage in both dams stood at 858,642ML. This is the last available data


from SEQWater. At this stage the River height at the dam was 73.77m.


• The river (and by assumption the Wivenhoe dam) continued to rise over the next six hours


and reached an interim peak of 74.51m at 14:57 Tue 11th.


• At this time the river level started to fall quickly to 74.19m around 17:45 Tue 11th


• Media reports indicate that the discharge from Wivenhoe was increased from 344,00ML/d


through 490,000ML/d (both reported by the Courier Mail) to an overnight peak of


645,000ML/d (reported in a media release by the Queensland Police Service).


• SEQWater reported that at 0730 on Wed 12th the releases from Wivenhoe were reduced


temporarily to 215,000ML/d to allow Lockyer Creek peak to enter Brisbane River and would


subsequently be increased to maintain a maximum flow through Moggill of 301,000ML/d.


Some additional data: -


• SEQWater report that there is a delay of approximately 36 hours between a release at


Wivenhoe and a peak at the Brisbane City Gauge.


• It is likely that the rapid drop in the river level at Wivenhoe commencing around 14:57 Tue


11th was due to a substantial increase in the discharge rate from Wivenhoe (645,000ML/d?).


• At around the same time, 16:03 on Tue 11th the Courier Mail reported that Wivenhoe was at


190% and Somerset at 176%, indicating a total capacity for additional flood storage of


636,000ML.


• 36 hours from 14:57 Tue 11th is 02:57 Thu 13th which corresponds almost identically with the


peak of 4.46m experienced at the Brisbane City Gauge.


Now for an attempt at interpretation of this sequence: -


• SEQWater were very slow to respond to the initial increase in levels at Wivenhoe and took 6


days before there was any real increase in rate of release from Wivenhoe to return the dam


to proper flood management levels. Even though there is apparently a legislated


requirement to manage this over 7 days.


• SEQWater then substantially over responded during the afternoon of Tue 11th increasing the


discharge to 645,000ML/d(?). This was at a much higher rate than the current water inflows,


resulting in a substantial drop in the level in Wivenhoe. This was even though there was


approximately 636,000ML of capacity available for additional flood storage in Wivenhoe and


Somerset.


• This substantial increase in the discharge from Wivenhoe was the cause (sole cause?) of the


peak in the Brisbane River.


• Early on Wed 12th (07:30), SEQWater recognised that this discharge rate was excessive and


reduced it substantially to 215,000ML/d. This discharge rate has been sufficient to ensure


that the river level at Wivenhoe (and presumably the dam) did not continue to increase and


indeed has allowed the level to gradually fall.


• Hindsight is a wonderful thing; but there are really two decisions that were taken which


would seem to have little justification at the time they were made. These are: -


o Not increasing releases from Wivenhoe between the 4th and the 10th when the dam


levels were rising and there was no downstream flooding, and


o The decision to substantially increase the discharge rate from Wivenhoe to a peak of


645,000ML/d on the afternoon/night Tue 11th.


• It is difficult to understand the justification for this increase in the discharge rate, especially


as the decision was substantially reversed within about 12 hrs. In addition even if SEQWater


had released at the low rate of 215,000ML/d (which is the rate it was reduced to at 07:30


Wed) rather than a maximum rate of 645,000ML/d for the whole time from 14:57 on Tues it


would have used only 306,000ML of the available free capacity in both dams of apparently


859,000ML.


The serious questions are: -


• Why did SEQWater not allow the total available flood storage capacity of Wivenhoe to be


utilised during this period?


• What justification was there for the substantial increase in discharge from Wivenhoe to


645,000ML/d when a release rate of 215,000ML/d has been demonstrably sufficient to stop


the levels in Wivenhoe rising and while there remained substantial capacity in Wivenhoe for


additional flood storage?


• Was this increase to 645,000ML/d the sole reason for the significant flooding in Brisbane?


• Why did it initially take SEQWater 6 days to respond to the gradually increasing water levels


in Wivenhoe which reduced its flood control capacity?