As part of the Brady Review into fatalities in the mining and quarrying industry, we examined the causes of serious accidents. These types of accidents have a specific definition in Queensland – they’re accidents that either result in a fatality or result in a person being admitted to hospital as an in-patient for treatment of their injury. Systematic reporting of these accidents to the regulator began in about 2012, and a total of 589 were considered in the analysis.

In this article we examine two aspects of these type of incidents. What caused them? And how did the industry respond to them in order to prevent reoccurrence?

What we found was a fundamental failure to effectively identify and control hazards, and once these incidents had occurred, we found the controls put in place to prevent reoccurrence were typically some of the least effective available.


As part of the reporting process to the regulator, the cause of each serious accident was categorised with respect to whether or not a hazard was identified, and if it was, was it adequately controlled. The various categories are shown below. 

This pie chart indicates that the causes of serious accidents were dominated by two categories. 

A total of 36% were caused by a failure to identify the hazard, while 45% were a result of a hazard being successfully identified, but not effectively controlled.

This means that more than 80% of serious accidents are a result of a very fundamental failure in hazard management.

And it’s worth stepping back here and talking briefly about the fatalities we examined as part of this review. A total of 47 occurred between the beginning of January 2000 and the end of July 2019, and in almost all of these incidents we found that ineffective controls also played a key role.

So not only do ineffective controls cause serious accidents, they also cause fatalities.

Response to serious accidents

As part of the reporting process, the corrective action put in place after these incidents was reported to the regulator. This was submitted in free text format, but as part of the review we grouped the reported controls into the various categories of the hierarchy of controls.

If you’re not familiar with the hierarchy, below is a representation, running from most effective to least effective control. 

The most effective controls are elimination, where you remove the hazard, followed by substitution where you substitute the hazard with something less hazardous, then isolation where you isolate the hazard from people, and then engineering controls, such as guards on machinery. These types of controls are known as hard controls because they physically stop the hazard from causing harm to people.

Then you have administrative controls, such as procedures and training, and finally there’s PPE (personal protective equipment), eg, safety boots and gloves. These final two controls are known as soft controls, and they control hazards in a very different way from hard controls. Rather than having an effect on the hazard itself, they’re attempting to have an effect on the people who interact with the hazard. 

But the key takeaway from this discussion is that there are different types of controls, and there’s a very clear hierarchy of their effectiveness.

So were the controls put in place in the aftermath of a serious accident some of the most effective type of controls available?

Well, the pie chart below shows the highest level of control applied in the aftermath of these accidents. For example, if two types of controls were reported for a single incident, say isolation and an administrative, only the highest level applied is shown. 

In 62% of the incidents the highest reported controls were administrative in nature. Further, less than 30% were hard controls. So following an incident that had a demonstrated ability to cause a fatality or a serious accident, one of the least effective types of controls was most commonly applied. In a very pessimistic sense, applying an administrative control is essentially telling people not to let it happen again.

We can also examine how the various types of controls are applied over each financial year. 

This data indicates that the 2016/17 financial year had the highest percentage of application of hard controls – elimination, isolation, substitution and engineering controls accounted for almost 50% of the controls applied. But from that year onwards there has been a decrease in the percentage of hard controls applied, dropping to less than 30% in the 2018/19 year.

One way of attempting to capture the typical words used in the reported controls was to represent the reported data, which was in free text format, as a word cloud. The words below are different sizes according to how often then appear in the text – the bigger the word, the more often it appears. 

As you can see the reported administrative controls are dominated with words such as review, ensure, and procedure.


So what does this all mean?

Well the first point is that accidents are happening because of a failure of fundamentals: a failure to identify a hazard and control it effectively.

The second point relates to the industry’s response to the serious accidents. From discussions with people in the industry, this was one of the most confronting, but not surprising, findings from the analysis.

It was confronting because it highlighted the extent of the industry’s reliance on administrative controls to manage hazards, but it wasn’t surprising because many in the industry told me that selecting a paperwork ‘solution’ was a default selection – the industry is simply heavily reliant on managing risk using paper.

And if the leading cause of serious accidents is ineffective controls, and the leading type of controls put in place in the aftermath of an incident are administrative controls – one of the least effective controls available – is it really that surprising that we then see ineffective controls playing such a key causative role in fatalities and serious accidents?

On the upside, this is a real opportunity for the industry to have an honest conversation with itself about its reliance on administrative controls, because moving towards selecting harder controls could have a significant impact on industry safety. 

To hear more check out episode 3 of Rethinking Safety or refer to Section 4.4 of the Brady Review into mining fatalities. All figures reproduced from the report or associated analysis.

Author: Sean Brady is the managing director of Brady Heywood (, based in Brisbane, Australia. The firm provides forensic and investigative structural engineering services and specialises in determining the cause of engineering failure and non-performance.