News is filtering out this week from the Queensland Flood inquiry. More than a year ago massive rain fall triggered flooding over most of the Queensland, Australia area. 35 were confirmed dead. A commission is looking into whether mismanagement at Wivenhoe Dam occurred during the flooding. Some of the statements from those they are interviewing are honest, so honest it surprises me. But it also strikes a chord with engineers I know and unfortunately sometimes the engineer in me as well.
The four engineers who controlled the dam are now being called to respond to conflicting evidence.
John Ruffini is the latest engineer to be called to respond to allegations of conflicting evidence.
The dam operator’s official account of events written after the floods says the engineers began focusing on protecting urban areas from flooding on Saturday, January 8.
But another report written during the event shows that did not start until two days later.
Mr Ruffini blamed that report on his poor choice of words.
“I’d been up for 12 hours, I’m an engineer and my English is crap,” he said.
I hear him. Sometimes I look back at something I wrote. Something that four more people had to sign their names to. And stuff doesn’t even make sense. Not that I’m excusing what happened, but boy can I sympathize.
Mr Callaghan also questioned whether Mr Ruffini had written down or discussed the strategies he used in the lead-up to last year’s flood.
Mr Ruffini explained he would have discussed them with the other engineers.
He was then pressed to explain which part of the strategies he would have referred to when making decisions.
The engineer stumbled through his responses at one point replying, “the spreadsheet stuff”.
Loves me my spreadsheets. And we engineers are not conditioned to sit on panels and answer questions for an inquiry. Though his non-specificity here might also indicate he lacked the real technical knowledge to make the call or to fully understand what kind of call he was making. Hard to tell. Then behind door #2:
An engineer has told the Queensland floods inquiry that keeping a contemporary record of water release strategies was “too time consuming”.
He said technical terms used for each water release strategy meant little, if anything, to anyone except the dam’s operators until after Brisbane and Ipswich flooded.
Mr Malone even questioned if the State’s Water Minister Stephen Robertson knew the difference between a W1 and a W3 release at the time.
Oh snap you stupid bureaucrats, you wouldn’t have understood it anyways. On the other hand, I totally get the whole negativity towards tracking and logging everything. At some point when you’re trying to log all the stuff you’re doing, it takes you longer than it took you to just close the task. Dam safety and risk analyst McDonald was asked to make an assessment from the engineers’ official 1,100 page report. After hearing new information about the report:
Counsel assisting the inquiry Elizabeth Wilson SC asked: “If you assume it was based on a reconstruction of the events, would that change your opinion?”
He replied: “It likely would, because according to which strategy you’re in there’s a priority of objectives and you need that mindset to be operating the dam.”
Mr McDonald was then shown a ministerial briefing note he had not had access to, which shows the engineers in a different strategy to their official report.
Upon seeing the note he commented that “it would have caused me to form the opinion that the operators have not complied with the manual”.
Not using the manual? No way! Sometimes a couple years into a process someone audits a written procedure and realizes we forgot to specify they re-use some bolts or something when they are replacing a part. A couple years in when someone finds this error I think to myself, well how did they do it for the last two years? Or when a part is being triggered for failure based on hitting a particular value I wonder, well didn’t this happen before? Usually it did. Usually the technician just re-uses the bolts when the manual doesn’t tell him to. Sometimes someone doesn’t pay attention to what should technically be a failure number because they know it’s intuitively not a failure value (say, when the device is not operating). I’m not condoning not following procedure. And maybe that’s a reason for why the faults occurred. But like with the tracking log, an engineer often does not have proper time to update a record. And one’s understanding of a manual needs to be on an intuitive level not on a procedural level.