Latest / Safe As Podcast / E53: Weak Controls. Repeat Deaths. The Brady review into mining fatalities

Transcript

E53: Weak Controls. Repeat Deaths. The Brady review into mining fatalities · Safe As Podcast

• 0:07
  Imagine a world where workplace fatalities and serious,
• 0:11
  irreversible injuries are addressed with administrative
• 0:14
  actions, typically the least effective approaches.
• 0:18
  Well, this isn't imaginative fiction, it's the world we live
• 0:22
  in. Good day everyone.
• 0:25
  I'm Ben Hutchinson and This is Safe as a podcast dedicated to
• 0:28
  the thrifty analysis of safety, risk and performance research.
• 0:33
  Visit safetyinsights.org for more research.
• 0:36
  Today's report is from Sean Brady, titled Review of All
• 0:40
  Fatal Accidents in Queensland Mines and quarries from 2000,
• 0:46
  2019. Published in 2019.
• 0:49
  It was a report commissioned by the Minister for Natural
• 0:52
  Resources to review all Queensland Mines Inquiry's
• 0:55
  fatalities from 2000 to 2019. Oddly enough for definitions, a
• 1:01
  serious accident is an accident resulting in a fatality or a
• 1:05
  person being admitted to hospital as an inpatient for
• 1:08
  treatment resulting from their workplace injury.
• 1:11
  There's way too much to cover in this report, so I'm focusing on
• 1:15
  a few key items, particularly around the functioning of
• 1:18
  controls. So what did they find?
• 1:21
  So just some background on the injury and fatality trends and
• 1:24
  distributions is that the industry sectors with the
• 1:29
  highest fatalities typically also had the most serious
• 1:33
  accidents. Except for underground coal, the
• 1:36
  number of serious accidents has generally trended upwards over
• 1:39
  the last five years, mirroring an increase in the hours worked
• 1:43
  over the past four years. They note that the serious
• 1:46
  accident frequency rate is consistently trending upwards,
• 1:50
  indicating the industry is becoming more harmful, and a
• 1:53
  higher percentage of workers are likely to require
• 1:56
  hospitalisation for injuries compared to the previous five
• 1:59
  years. And the serious accident
• 2:01
  frequency rate for contractors is higher than for employees.
• 2:06
  They both are generally trending upwards and reasonably
• 2:09
  comparable, which doesn't necessarily support the view
• 2:13
  that employees work in a considerably safer manner than
• 2:15
  contractors or work in a considerably safer environment.
• 2:20
  More interesting, I think, is the functionality of the
• 2:23
  barriers of the controls and also the corrective actions that
• 2:26
  resulted from these serious incidents and fatalities.
• 2:30
  Most telling is that analysis of the corrective actions
• 2:33
  implemented after the serious accidents reveals that 62% were
• 2:38
  administrative controls. In contrast, less than 30% of
• 2:42
  the controls that were applied were hard controls, as in
• 2:47
  elimination, substitution, isolation or engineering
• 2:50
  controls. This is concerning because
• 2:53
  administrative controls, things like procedures and signs, et
• 2:56
  cetera, are among the least effective controls available
• 3:00
  within the hierarchy of controls, yet they are
• 3:02
  predominantly used to address hazards that have caused deaths
• 3:06
  or required hospitalisation. If this reporting accurately
• 3:09
  reflects the industry's actual responses, it implies that
• 3:13
  opportunities to effectively control hazards are being
• 3:16
  missed, leaving them within the system and exposed to workers.
• 3:21
  They're relying on people to have to ultimately respond to
• 3:24
  these hazards in practise. Specifically, they found that
• 3:28
  the percentage of high controls that were applied peaked around
• 3:31
  50% in 2016 two 1017. But largely administrative
• 3:37
  controls remain the largest category consistently applied.
• 3:41
  They look at this concept of absent or fail controls, finding
• 3:45
  approximately 81% of the serious accidents occurred in a
• 3:49
  situation where a hazard wasn't identified or the controls that
• 3:53
  were applied were ineffective, and the number of serious
• 3:57
  accidents with ineffective controls or hazards that weren't
• 4:00
  identified was found to be increasing within recent years.
• 4:06
  Interestingly, bypassed or unenforced controls based on
• 4:10
  this data were less likely to play a role within their serious
• 4:13
  accidents. Next, they unpacked high
• 4:16
  potential incidents. A high potential incident or a
• 4:19
  hypo is an event or series of events that has the potential to
• 4:24
  result in a significant adverse effect on the safety or health
• 4:27
  of people. This review included over 34,000
• 4:32
  hypos. Most of the hypos up to 85%
• 4:35
  didn't result in injuries. These are considered near misses
• 4:39
  or near hits and how you prefer the term.
• 4:42
  They're said to offer valuable opportunities to identify and
• 4:44
  remove hazards before they result in harm.
• 4:47
  A greater number of the hypos were reported in open cut coal
• 4:51
  consistent with the high work hours work in that sector.
• 4:55
  The hyper frequency rates, which is the number of hypos per
• 4:58
  million hours worked has been relatively constant for the last
• 5:02
  15 years at approximately 23 hypos per million hours worked.
• 5:08
  This constant rate, despite increasing hours worked,
• 5:11
  suggests that hazards are present and waiting for
• 5:14
  additional people to be exposed to them rather than being
• 5:17
  effectively removed or controlled at the source.
• 5:20
  And the reported rate of 1.4 Hbis or hypos per person per 30
• 5:25
  year career appears inconsistent with the actual experience of
• 5:29
  workers, suggesting significant underreporting of the hypos.
• 5:34
  Now for the corrective actions after the hypos, similar to the
• 5:38
  serious accidents, the percentage of high controls
• 5:40
  applied after the hypers is relatively low, around 25 to
• 5:44
  30%, while administrative controls constitute the largest
• 5:48
  category at 49% and even up to 60% in some years.
• 5:53
  Again, this heavy reliance on administrative controls suggests
• 5:56
  that the industry is missing opportunities to effectively
• 5:58
  control the hazards, and it seems to be a default tendency
• 6:02
  to apply more administrative controls like procedures, rather
• 6:06
  than stepping through the hierarchy of controls.
• 6:09
  But regardless of the reason, it suggests that hazards aren't
• 6:12
  being removed or effectively controlled at the source and are
• 6:16
  likely to remain within the industry.
• 6:18
  Similarly, again, when they look at absent or failed controls, a
• 6:21
  significant number of the hypers resulting from hazards that
• 6:25
  weren't identified and ineffective controls.
• 6:30
  Next, they look at lost time injuries.
• 6:31
  I've skipped most of this, but I'll just cover a couple of
• 6:34
  points. One is that the LTIF are the
• 6:37
  lost time. Injury frequency rate for
• 6:39
  contractors is on average lower than for employees,
• 6:43
  approximately .7 times lower. It's hypothesised that this
• 6:48
  results from under reporting by contractors, possibly
• 6:51
  incentivized by contractual agreements.
• 6:54
  Now they essentially win and lose work based on the stats
• 6:57
  that they present, and injuries to employees show considerably
• 7:01
  higher numbers for hidden injuries like back, knee,
• 7:04
  shoulder and neck complaints, which are less likely to be
• 7:07
  noticed if unreported. Unlike more sort of obvious
• 7:10
  injuries and similar to the serious accidents and the hypos,
• 7:15
  the industry response to the LTIS is overwhelmingly
• 7:18
  administrative in nature, accounting for 61% of corrective
• 7:22
  actions, and the application of administrative controls has
• 7:26
  remained consistently high over time.
• 7:29
  So what do we make of the findings?
• 7:32
  A couple of practical implications is considering what
• 7:34
  improvements flow from prospective hazard analysis and
• 7:38
  risk assessment, including deeper learning from daily work
• 7:42
  evaluating constraints, variability, resources,
• 7:45
  production pressures, et cetera. And 2nd, considering safety in
• 7:51
  design and the efforts to eliminate or mitigate hazardous
• 7:54
  energy interactions with people. Now this report was pretty
• 7:59
  scathing around the focus on administrative controls after
• 8:02
  serious, even fatal accidents. But there's also a counter
• 8:07
  response from some people that maybe higher order controls for
• 8:10
  some of these just aren't reasonably practicable.
• 8:12
  Maybe they've done as much as they reasonably can.
• 8:15
  And the last remaining bastion of controls, risk controls,
• 8:18
  really lie in the administrative domain.
• 8:20
  I don't know if the evidence supports that, but that is a
• 8:22
  belief by some. In any case, while the hierarchy
• 8:26
  of control is required or recommended in many
• 8:29
  jurisdictions around the world, I think Haddon Junior's work is
• 8:33
  more illuminating. Instead of the hierarchy of
• 8:36
  controls, Haddon suggested 10 countermeasures.
• 8:40
  The first is to prevent the initial marshalling of the form
• 8:42
  of energy. This really aligns with
• 8:44
  elimination. To reduce the amount of energy
• 8:47
  marshalled. To prevent the release of the
• 8:50
  energy. To modify the rates of the
• 8:53
  spatial distribution of release of energy from its source.
• 8:57
  So the overall area to separate in space or time the energy
• 9:00
  being released from between the agent and the energy.
• 9:04
  To separate the energy being released from the susceptible
• 9:07
  agent by placing in some sort of material barrier.
• 9:11
  To modify the contact surface or the subsurface or the basic
• 9:14
  structure which can be impacted, to make it more robust and
• 9:17
  resilient, to strengthen the structure which might be damaged
• 9:20
  by the energy transfer, to move rapidly in detection and
• 9:25
  evaluation of damage and to counter its continuation and
• 9:28
  extension. So how we respond to the
• 9:30
  damaging energy and all those measures which fall between the
• 9:34
  emergency period following the damaging energy exchange and the
• 9:37
  final stabilisation of the process.
• 9:40
  I know that was a bit technical, but really I do recommend you
• 9:43
  search do a Google search for Haddon's 10 countermeasures.
• 9:47
  Probably provide quite a new perspective around how we can
• 9:50
  intervene in the exchange of hazardous energy.
• 9:54
  Finally, the report really nicely challenges some injury
• 9:58
  measures like Ltis by pointing at their limitations.
• 10:02
  The LTI frequency rate. LTIFR is considered a poor
• 10:06
  measure for monitoring safety capacities.
• 10:09
  Ltis are prone to manipulation as they measure how injuries are
• 10:13
  managed after they occur, not how safely an organisation is
• 10:17
  performing. It's possible to reduce the
• 10:20
  LTIFR without actually getting better at managing hazardous
• 10:24
  energy exposures. Many underlying factors for
• 10:27
  fatalities would not have caused injuries prior to the fatality,
• 10:31
  meaning they wouldn't be recorded as Ltis and thus remain
• 10:34
  unidentified as issues. And really aptly, according to
• 10:39
  the report, LTIFR is a distraction and at worst, it
• 10:44
  causes early warning signs to be missed.
• 10:47
  And just remember that a limitation, or at least a nuance
• 10:50
  with these types of reports based on investigations is that
• 10:54
  they're not clear cut black and white explorations of facts and
• 10:58
  the objective world. Instead, investigations are
• 11:00
  rather constructed narratives by people with their own
• 11:03
  subjectivities, worldviews, biases, et cetera.
• 11:07
  So investigations are more constructions than simply
• 11:10
  collections of objective facts. That's it.
• 11:14
  On Safe as I'm Ben Hutchinson, please help share, rate and
• 11:18
  review and checkoutsafetyinsights.org for
• 11:21
  more research. Finally, feel free to support
• 11:24
  Safe As by shouting a coffee link in the show notes.