Sunday, March 18, 2012

Thomas, the Australian, and the "avoidable flood" that was unavoidable

Hedley Thomas is, predictably, crowing about the finding of the Queensland Flood Inquiry that the Wivenhoe Dam was not operated in accordance with the manual. It's true, his articles did mean the commission had to examine how the SEQWater report came to be written because (it would appear) lawyers missed the importance of some evidence relating to what dam release strategy the engineers thought they were operating on during the weekend of 8 - 9 January.

I haven't read all of the Inquiry's report on the engineer's evidence; but certainly, the way the SEQWater report was written was (as I indicted in a previous post), not a good look.

Whether the 3 engineers in question deserve any condemnation for it from a Crimes and Misconduct Commission enquiry is another question. The Inquiry found that the the Manual was pretty hopeless anyway, and I suspect that the fundamental problem with any such manual is that, by its very nature, flood management via dam releases must leave a wide scope of discretion in the engineers, at least until you get to a level where it's a case of "do anything to protect the dam."

And at the end of the day, what should matter is the degree to which, if any, dam operations actually caused any harm.

This is where Hedley Thomas and the Australian deserve no praise at all; quite the opposite in fact. The paper as a whole has followed a sensationalist line in its reporting, virtually since the flood occurred.

For example: this is what the Inquiry writes at Chapter 16, page 527:

It is unfortunate that there has been a conflation in some media reporting of two separate issues: whether there was non-compliance with the manual strategies and whether it caused unnecessary flooding. The Commission has found the first (see 16.11 Conclusions: the dam operations strategies.) As to the second, Mr Babister’s perception was that the flood engineers managed Wivenhoe Dam so that its flood mitigation effect was ‘very close’ to the maximum achievable within the constraints of the manual. That may well be right. The problem is that the possibility exists that because the engineers failed to consider the releases open to them within the parameters of the correct W strategy, an opportunity may have been lost for earlier releases.

The evidence was uniformly to the effect that the pattern of releases adopted on Saturday 8 January was appropriate: the lake level was only just over 68.5 metres and showed every sign of dropping; higher releases would have been risky and unwarranted. The picture is not so clear for Sunday 9 January, when the rainfall returned.
Was that opening line directed to Hedley Thomas? I think it pretty likely, given that this is what we get from Hedley Thomas (and Jamie Walker):
The Floods Commission of Inquiry's finding that the engineers who operated Australia's largest dam failed to adopt the correct strategy to protect Brisbane from inundation for about 36 hours from Saturday, January 8, last year, has given a major boost to the hopes of thousands of victims.
See the difference?

The Inquiry notes that, even though the dam manual was not followed from 8 am Saturday, no one thought the actual dam releases were at inappropriate levels during Saturday. The following paragraphs (at page 527 of the inquiry report) indicate that it was probably during the afternoon of Sunday 9 January that the change to faster releases could have happened:

Mr Babister initially said that ‘the more practical or realistic options if you were going to have higher releases, is to start some time after midday or somewhere between midday and 1600 hours. That’s when it would be realistic on the 9th to increase flows above what was released’; although he subsequently modified that view to say that the ‘only area’ that there was ‘some argument they probably could have released slightly higher flows’ was after 4.00 pm that afternoon. The scenario of higher releases on the afternoon of 9 January, Mr Babister said, was most closely reflected in scenario 9 of Figure 16.1; but it was ‘an adventurous risk-taking approach’ because it relied on confidence in the rainfall forecast.

Mr Shannon’s view was that given the ‘frightening’ inflow by 2.00 pm on 9 January and the predicted lake level it would be ‘extraordinary’ not to have put the closure of the bridges in train by then, in accordance with the intention of W3. And Mr Tibaldi volunteered in evidence that ‘decid[ing] to ramp up earlier for this event... would have reduced flood damage’. Mr Ayre agreed.
The chapter concludes:
That night [9 January], though, at about 7.00 pm, it was recognised that the release rate from Wivenhoe would have to be elevated. No actual strategy change was documented; at best, it can be said that the actions taken were consistent with strategy W3.

It follows that Wivenhoe Dam was operated in breach of the manual from 8.00 am on 8 January 2011 until the evening of 9 January 2011.
It therefore seems accurate to say that the Inquiry has only raised doubt about the actual rates of water releases for only 7 hours (from midday to 7 pm on 9 January) or even less.

Does anyone really think that 7 hours of faster release would have made a huge difference?

Does anyone reading the Australian or Hedley Thomas get any sense of that?

The Inquiry finds:
There is, it is obvious, plenty of scope for argument about whether adherence to the manual strategies would have made a difference to the way in which the flood engineers actually operated the dam; but the possibility certainly exists that they would have responded more quickly to the developing conditions of 9 January had their mindset been one of applying strategy W3. Ascertaining the practical result of acting more quickly also is subject to the uncertainties inherent in the modelling; but again, the possibility exists of at least some improvement in the flooding outcome for Brisbane and Ipswich.
Here's how Thomas interprets this:
Supreme Court of Appeal judge Catherine Holmes SC found that "the possibility exists of at least some improvement in the flooding outcome for Brisbane and Ipswich" if the dam had not been mismanaged. This is a departure from earlier findings made by the inquiry's expert witness, hydrologist Mark Babister, that the flood engineers had achieved close to the best possible result in mitigating the flood.
I don't see how it is a departure at all. As I understand it, Babister did not change his advice to the Inquiry between his two appearance - his modelling on different scenarios indicates that, for large parts of Brisbane, the flood might have been capable of being reduced by 30cm to a 90 cm (see page 526 of the report). The Inquiry notes the modelling has considerable uncertainties, and the scenarios they asked Mr Babister to model are not even all "realistic".

Of crucial importance is this paragraph - talking about what would have happened even if you started with the dam at 75% capacity:
It is important to note that even at these lower river heights, major flooding would still have been experienced in Brisbane. The Bureau of Meteorology defines a major flood as one which peaks above 15.5 metres at Moggill and 3.5 metres at Brisbane city1028 (the Port Office gauge). Scenario 4, which involved an initial lake level of 75 per cent of full supply level and W strategy trigger levels reduced by 25 per cent, resulted in a modelled height of 16.3 metres at Moggill and 4.0 metres at the Port Office.
To remind you,the measured height of this flood at the Port Office was 4.46m (although another gauge indicated only 4.27 m - see page 522 of the inquiry report).

So let's get this clear - the modelling of the independent hydrologist, based on starting at a dam 75% and with lower "trigger points" for releases would have resulted in a flood in Brisbane city of about 50 centimetres less.

Contrast this to what Hedley Thomas was writing on Feb 14 2011:
THE clearest official acknowledgment that the devastating flood in the Brisbane River was avoidable has been the decision yesterday to let go 25 per cent of the water stored in the Wivenhoe Dam.
Sorry, Hedley, Inquiry says "no".

Or what about the headlines given to his continual promoting of the idea that the dam management caused the flood:

  • The great avoidable flood: an inquiry's challenge [22 Jan 2011]
  • Engineer bores a hole in dam untruths [19 March 2011 - a piece promoting the engineer Michael O'Brien's figurings given the title "Brisbane Flooding January 2011: An Avoidable Disaster".]
And now:
  • Damages to flow from Wivenhoe Dam breach [17 March 2012] even though the body of the report contains the caution from Maurice Blackburn lawyers: "If the action proceeds, it is likely to be the largest class action Australia has ever seen."
Thomas claims yesterday in a separate article:
The inquiry's expert witness had previously asserted that close to the best possible result was achieved; however, independent engineers consulted by The Australian have calculated that almost all of the flooding could have been avoided.

The detailed modelling necessary to determine this will be conducted by overseas experts engaged by law firms Maurice Blackburn and Slater & Gordon, which yesterday described the finding of the breach and the cover-up as "crystal-clear".
Well those "independent engineers" obviously aren't good enough for the Australian based litigation lawyers. In fact, if you look at Michael O'Brien's report, which Thomas was promoting in the report note above, O'Brien's work experience has been in building gas and oil pipelines, but he has had "to assess the impact of various rainfall events and to interpret and rely on flood mapping for the design and location of process facilities." Colour me unimpressed. His entire paper appears to be a mere series of "what ifs" in terms of when water might have been released if you had perfect knowledge of the rainfall that would arrive in the next few days, and is not (as far as I can tell) based on hydrological modelling at all.

Here's the thing: Hedley Thomas decided early to go hard with the story that this was a "preventable flood" that was the fault of dam operations. This was based on some hunches of a couple of engineers, and Thomas and the Australian has, in a long series of headlines and articles continued to foster this belief.

In reality, the Inquiry and the independent modelling it used has shown it was not an "avoidable flood" at all. Different timing of water releases may have made a relatively small difference to flood levels to most areas, but it still would have been a major flood even if you started at a 75% dam level and had lower triggers.

Given this scenario, my hunch is that it is rather unlikely that overseas modelling is going to be certain enough to allow for liability to be legally established for anyone. Certainly, the inquiry modelling would indicate that no one (in most of Brisbane, anyway) with more than about .5 m of water through their house is going to have any hope of blaming their damage on dam management.

The Australian, and Hedley Thomas, have been largely uninterested in reporting this level of detail of the Inquiry and its modelling, and have been more interested in campaigning for a interpretation of the event that actually isn't holding up to scrutiny. Personally, I think the the misunderstanding in large parts of the community that they have fostered for a year about the nature of the flood easily outweighs any benefit of having successfully made a few engineer's lives a further misery.

Finally, it's interesting too to note the connection between climate change skepticism (for which The Australian is well for promoting) and the "avoidable flood" meme. On both subjects, people like Andrew Bolt have been happy to promote the Thomas line without actually looking at the detail of the Inquiry. Same as his ignoring the fact that more intense droughts and floods have been predicted by CSIRO for years, Andrew Bolt has shown no sign of informing himself of the Inquiry's detailed findings as illustrated in this post.

Increasingly, I have been noticing how "pop" climate change skepticism thrives on laziness, and not looking into matters in enough detail. These "fake skeptics", as some call them, are easily conned in all sorts of ways, and The Australian is always there to help.

1 comment:

Anonymous said...

I wrote the following on another story which is to the same effect. It includes some back-of-the-envelope calculations that pretty much confirm that the mismanagement via incorrect W strategies had a trivial effect on flooding:

---

"""This shows the hydrograph at Wivenhoe during the flood event:

http://wivenhoesomersetrainfall.com/images/2011%20Major%20Minor%202011_0001.jpg

The darker blue is inflows which is an independent variable in this case. The others are dependents - the light blue represents outflow releases and red represents headwater elevation (which rises when inflows > outflows, and falls when outflows > inflows... simple enough!).

Also, the flood report has been released which contends that the engineers did breach the manual and were operating the dam in W1 up to Sunday evening instead of moving to W3 on Saturday morning. Looking at the hydrograph the information present at that time (8th to 9th) show that the catastrophic inflows that caused the headwater elevation to grow beyond ~115% capacity occurred on the cusp of their transition to W2/W3 (which is represented by the indented rise in outflows) on the evening of Sunday 9 January, and then to W4 (the huge increase) on Tuesday when the headwater elevation reached beyond 74m. Headwater elevation was stabilising since inflows were close to outflows around Monday night/Tuesday morning, before the second great pulse of inflows which made it necessary to move to W4 and drastically ramp up releases. Based on the data alone, they were in breach of the manual as W2/W3 should have been initiated earlier (the dam reached the transition point 68.5m on the weekend but they didn't move to higher releases until Sunday evening), but based on flow rates downstream ("protecting urban areas from flooding" mandate in W2/W3 is based on a "threshold for damaging floods" of 4,000m3/s at Moggill) they could have boosted releases by about 1,000m3/s or 84,600ML/day (flow rates were ~2,900m3/s at Moggill for the week beforehand, so they only had 1,000m3/s of releases that would adequately protect urban areas from flooding). Over 36 hours this would have contributed to an extra release of 126,900ML or a pitiful 5.4% capacity of the dam. At the inflow rates later on Sunday, after their "breach of manual" period as outlined in the Commission Report, that extra capacity would have bought the engineers about 30 minutes reprieve. This analysis basically concludes that their breach of the manual led to a trivial change in flood mitigation outcome - indeed, the modelling done for the commission by Sinclair Knight-Mertz shows that under an ideal perfect scenario the flood could have been reduced by 90cm at Moggill and 30cm at the City Gauge. At the peak height of the flood, Wivenhoe's releases contributed only half the total volumes; the rest was Lockyer, Bremer and then the sum of local catchments around Brisbane. Wivenhoe releases contributed the bulk of flood waters after the peak during the drawdown of the flood compartment.

The engineers rightly should stand at the CMC but it has nothing to do with mismanaging the dam: it is for misleading the commission with false information. Any claims of compensation in class action have to demonstrate that the damages were caused by mismanagement of the dam, inferring that if the dam was managed correctly (in this case, within the manual) the damages would not have happened. It is clear that this condition won't be met. This is intuitively just as well - you cannot blame Seqwater for damages that were going to happen regardless. I don't want to see people waste money paying expensive lawyers to sue Seqwater for the next 10 years to no effect.

---

Happy reading! (If you feel like trolling through my bad grammar).