This report, part of creative sentencing research, explored Serious Incidents and Fatalities (SIFs) in mining, and the causes, and the most effective controls for SIFs.
Another aim was around the fallibility of people, and when they make mistakes, ensuring there are adequate capacities “so that they ‘fail safely’”, rather than “rather than ‘failing lucky”, or worse, “failing unsafely/unlucky”.
See comments for a link to a summary of the preliminary report here.
Only a few extracts – it’s a large report and worth your time to read in full, so check the link in comments.
The methods involved three complimentary activities:
1) Workshops (200+ subject matter experts across 60+ companies)
2) Analysis of data pertaining to Critical Controls Assurance (CCA) vs SIF potential + actual events
3) Surveys and interviews
Results
For respondent perception data, people strongly aligned with the importance of social-psychological factors in shaping safety.
For instance, “ one of the strongest predictors of safety perceptions were crews who have positive safety attitude”, like crews placing strong emphasis on safety, or beliefs that they can undertake the work in a safe manner.
For reasons why hazards may not be identified or reported included novice or developing workforce, perceived career implications, and normalisation of high hazard work.
15% of respondents had a fear of career repercussions, and 11% feared social repercussions for reporting hazards.
Usually fewer, ~3-17% of people, feared specific hazards. Hence, in this sample, ‘fear’ of identifying or reporting hazards resulted more from social factors than technical.
Remember – I’ve skipped A LOT of findings, but they provide the following recommendations based on perception data:
1. Focusing on high-energy hazards / The Stuff that Kills You (STKY), and developing layered controls and ensuring the controls are in place and functional
2. Using visualisations to show the hazards and their control systems – e.g. bowties, critical task infographics, start work check diagrams and more
3. Share the visualisations with workers so they can “better see, monitor, and manage their workplace hazards, as the work is done (versus work as planned/imagined)”
4. Verify the efficacy of critical controls by comparing to SIFs, hazards and controls
5. Compare “workers’ perceptions of risk versus the hazards they are exposed to”, e.g. fear of reporting can drive underreporting, while over-confidence due to risk normalisation can result in hazardous practices
They also draw distinction between quality assurance, like demonstrated competencies, like working on a scaffold vs quality control, e.g. auditing the efficacy of barrier systems. This focus is said to “focus attention on how “work is done” in the field versus “work as planned”.
There is also an interaction effect, like a supervisor’s experience and competencies can improve the quality of field hazard analyses, which then create positive feedback loops: e.g. “if supervisors have demonstrated competency in working on scaffolds, they should recognize hazards associated with that work, ensure that controls are in place and functional, and create high- quality [hazard assessments]”.
Critical Control Performance and SIFs
Next they compare control performance expectations vs actual and potential incident data.
Importantly, “SIFs have complex precursors that are not easily visible in companies’ physical environment, reports, or what workers say”. Hence, many SIF precursors are only visible “watching the ‘work as done’ and how it differs from ‘work as intended”
And this applies even to routine work, where “controls may be insufficient, bypassed, or non functional”.
They believe that bowties are useful for visualising the presence of safety: “Bowties also illustrate that safety is the presence of layered controls rather than an absence of serious consequences”.
Nicely, they say that bowties “make the invisible visible and the unspeakable discussable”.
Further, the more layers of control employed, the more resilient is the safety system [** barring issues around greater complexity from redundancy in particular configurations, of course.]
With more resilient capacities, the system is “more likely to fail safe than fail lucky”.
They also talk about the hierarchy of control. Eliminating work at heights eliminates the risk of people falling to their death. Substituting non-toxics for toxic chemicals removes one harm potential is a pipe bursts. Lower order controls, relying on procedures or PPE, like work permits, in contrast, “are less effective [and] easy to bypass”.
Bowties also help visualise barrier/control degradation factors, like human and organisational factors, which support or degrade barrier/control performance.
Critical Control Assurance Activities
A really interesting part of their research compared actual/potential incident data vs critical control assurance activities (CCA; e.g. field verifications of control performance).
WAY too much to cover here, so just a few cherrypicked examples of their data.
First, some definitions.
Critical Control Assurance (CCA) is said to be a “promising method to prevent SIF by focusing on the high-energy, STCKY (stuff that can kill you) hazards”. In short, it’s an activity than confirms that critical controls are present as intended and “ensuring that we can fail safely”.
Drawing on the ICMM’s definitions, a control is an act, object or system that can independently prevent a threat or mitigate the consequences.
A critical control is a control where the absence or failure would significantly increase risk.
Mismatches between SIFs and Control Assurance Activities
Importantly, whereas dropped objects constitute between 50-80% of hazards in operational areas, CCA inspections only focused on these hazards in 37-45% of cases.
Worse still, whereas working with or near energised equipment or exposure to hazardous atmospheres constituted 5-60% of hazards, just 0.1 to 8% of CCAs focused on these hazards.
Even worse still, just 0.13% to 1% of CCAs focused on these former two hazards in other operational areas.
Therefore, “By identifying these mismatches between [potential and actual SIFs] and the attention of CCAs to prevent these, these preliminary findings help identify the potential for enhancements to the quality and attention of CCAs in this industry”.
Further, while confined spaces accounted for 17% of the SIF incidents, only 10% of CCAs captured these issues.
Working with or near energised equipment accounted for 25% of SIF incidents but “were captured in only 1% of CCAs”.
In other operational areas: For hazardous atmospheres or chemicals – whereas 6% of incidents related to these hazards just 0.13% of CCAs focus on those issues.
For electrical high voltage – whereas just 1% of SIFs related to HV, “almost no CCAs (0.01%) identified this indicator.”
The table below highlights more of these mismatches – remember I’ve skipped heaps of this stuff:
Recommendations:
Overall, they draw on Conklin’s HOP principles, being:
· People make mistakes
· Blame fixes nothing
· Context drives behaviour
· Learning is vital
· Response matters
Explaining the importance is that our systems should increase the chances of failing safely rather than failing lucky or failing unsafely/unlucky.
For understanding our SIF hazards and interactions, they offer some preliminary suggestions (directly quoting the paper):
· Create a hazard inventory of ‘work as done’ in all mine operations (mining, extraction, treatment, and tailings) and ancillary activities.
· Identify/develop indicators for SIF precursors, controls, and assurances/audits.
· Determine which activities are most likely to be completed by contractors.
· Examine how indicators, controls, and control/barrier assurance vary across owners/operators, contractors, and subcontractors.
· Survey the prevailing safety culture within owners, contractors, and subcontractors.
· Recommend systems-level, multi-layered, effective, and assured controls.
· Promote and support the implementation of these controls into owners/operators’, contractors’, and subcontractors’ safety management systems.
Shout me a coffee (one-off or monthly recurring)
Refs –
Final report: https://ualberta.scholaris.ca/server/api/core/bitstreams/59a0eca0-fcfc-4629-b6a0-c6ed0288ed51/content
LinkedIn post: https://www.linkedin.com/pulse/building-resilience-safety-management-systems-reduce-sifs-hutchinson-0x12c
SafetyInsights.org 