This website uses cookies. By continuing to browse the site you agree to our use of cookies. Find out more about the cookies we use and how to manage them.

Close

EEJ: Inside an Engineering Enterprise

Safety - time for a new direction

The UK construction industry operates some of the most sophisticated safety management procedures in the world – and is one of the most regulated.

Nevertheless, even on the most compliant of sites, the risk of fatal injury remains a reality.

A progressive new strategy, that has been successfully trialled in Australia, could provide the solution. Counterintuitive, on the face of it, it’s already showing early signs of success.


Safety in construction is at a crossroads. Despite a plethora of rules and regulations, systems and processes, the industry continues to cause harm. On average 44 people were killed on British sites every year between 2009 and 2014 – more than any other sector.
While its workforce makes up just 5 per cent of the national total, construction accounts for 31 per cent of employee fatalities and 10 per cent of reported major/specified injuries.

Saddening though they are, these statistics represent a significant improvement on previous decades. Even up until the 1990s, annual fatality rates averaged at 100-plus. However, over the past five years, the pace of progress has slowed. In that time, the number of deaths has remained fairly steady. If, as it would appear, we have reached a standstill, then clearly more of the same is not the answer.

Laing O’Rourke’s safety record 

As one of the UK’s largest contractors, Laing O’Rourke’s safety performance has inevitably followed the same trends as the rest of the sector. The company’s Disabling Incident Frequency Rate (DIFR) – which encompasses any incident resulting in an absence from work of one or more shifts – has remained at roughly 0.25 in the UK for the past five years. This is around five on a six-sigma scale, and among the best in the industry.

In 2010, the business launched Mission Zero – a global campaign aimed at eliminating all harm from its operations. Linked to this were two key targets: a 0.1 DIFR by 2015 and a 0.1 All Accident Frequency Rate (AAFR) by 2020. In 2015, the Group recorded a DIFR of 0.28 – an improvement on previous figures but nevertheless short of target.

Interestingly, however, while the UK performance plateaued during this period, the Australian business improved dramatically, with its DIFR dropping from 1.68 in 2010 to 0.17 in 2015.

How did this come to be – and what can we learn from it?

The Australian experience

Just five years ago, the Australian construction industry looked very much like the British construction industry of the 1970s and 1980s: mistrust between management and the workforce, exacerbated by often-antagonistic union interventions; adversarial contractual relationships; inflexible attitudes on all sides and high levels of accidents (though relatively few deaths).

Shortly after launching Mission Zero, Laing O’Rourke realised it was never going to achieve its targets without addressing the safety issues inherent within its burgeoning Australian business. In 2011, the company sent John Green, then Health, Safety and Environment Director for the Europe Hub, to Australia. His task: to introduce the tried-and-tested approach (a combination of systems, processes, training and culture) that had proved so successful in the UK.

Interestingly, while Australian architects and consulting engineers have a duty to provide a design of reasonable practicability, there is still no equivalent of the EU-driven CDM Regulations. The Australian Government monitored the impact of CDM from its launch in 2007 and decided there were insufficient benefits to replicating them locally.

Despite the absence of CDM, Green was able to get the Australia Hub accident rate down to a similar level of that in the UK within just three years. However, as had happened in the UK, the results started to plateau at around 0.25 DIFR.

Nevertheless, the business had proved it could not only adapt to change, but embrace it. The concept of ‘behavioural safety’ had been virtually unknown – and its introduction heralded a seismic shift in attitudes.

With that in mind, it was decided that the Australia Hub would be an appropriate testing ground for some of the more innovative ideas in health and safety that were beginning to gain currency. One of the main exponents of these new approaches is Sidney Dekker, a professor at Griffith University in Queensland.

Dekker was previously professor at Lund University, where he founded the Leonardo da Vinci Laboratory for Complexity Systems Thinking. He graduated in psychology in Holland and gained a doctorate in cognitive systems engineering from Ohio State University. His work on human error and safety has been recognised worldwide, and his ideas are neatly summarised in his peer-reviewed paper entitled: ‘Employees: a problem to control or solution to harness?’, which appeared in the American Society of Safety Engineers journal, Professional Safety, in August 2014.

In it he argues that safety interventions can either target people or the technological and organisational environment people work in. ‘The choice means either fitting people into fixed systems or engineering those systems so they are fit for people.’ Certainly this is not new. He cites US Army Air Force pilots during World War Two who, regardless of their experience and skill, reported errors in using cockpit controls. Instead of trying to change the pilots, the engineers realised they could – and should – change the design of the controls to reduce the risk of incidents. ‘Humans were no longer seen as just the cause of trouble; they were also the recipients of trouble,’ says Dekker.

‘However, most industries still seem to favour safety interventions that consider people as a problem to control, through procedures, compliance, standardisation and sanctions. This also means that safety is measured mainly by the absence of negatives,’ he says.

Dekker is particularly critical of having a ‘zero’ vision for safety. ‘A zero vision can encourage the suppression, discouragement or recategorisation of incident or injury data and lead to other ‘numbers games’ such as an inappropriate use of modified duties or return-to-work programmes.’ He cites a 1998 study of the Finnish construction and manufacturing sector, which showed sites reporting the fewest incidents had the highest fatality rates. ‘Low incident reporting rates might suggest workplaces where superiors are not open to hearing bad news, which might explain why those that report fewer incidents also suffer more fatal incidents.’

He is also critical of a safety culture which celebrates the absence of minor accidents while losing sight of the bigger picture. He cites the Deepwater Horizon oil rig, where the day before it blew up killing 11 people and causing the world’s worst environmental disaster, senior executives from BP and Transoceanic were on board handing out awards for long periods of accident free operation. They failed to notice during their safety tour that the rig was operating outside its normal safe limits. ‘While measurable safety successes were celebrated, the organisation’s coherent understanding of engineering risk across a complex network of contractors had apparently eroded,’ says Dekker.

Dekker’s alternative approach to safety, and one which is gaining a groundswell of support around the world – despite much hand-wringing from the safety profession – is that ‘safety professionals should view people as a solution to harness rather than a problem to control. They should consider safety more as a presence of positive capacities rather than the absence of negative accidents. And they should move from a vocabulary of control, constraint and deficit, into one of empowerment, diversity and human opportunity’.

Green knew a different approach was needed if the rate of serious injuries was to reduce any further. So he met with Dekker in 2013 to discuss how his somewhat disruptive theory could be trialled in the context of construction. 

Trialling an innovative strategy 

An analysis of Laing O’Rourke’s existing safety strategy showed that it conformed to Dekker’s ‘old’ paradigms: people are a problem that must be controlled (the system is perfect, people are fallible); safety is defined (and measured) as the absence of negatives (not the presence of positives); safety management is designed to protect upwards, rather than downwards (its primary function is to create the ‘necessary’ paper trail).

Dekker’s advice was to examine things from the other way round, starting by looking at people as the solution. It is these people after all who are responsible for the 99,998 hours out of 100,000 that pass without incident. In every one of those hours they plug the gaps in the process and complete their tasks using their expertise, intuition and innate desire to be safe. It is these positive factors that should be measured, along with people’s resilience to accidents and their capacity to adapt and recover.

He also said safety should be regarded as an ethical responsibility, focused on protecting people from harm – as opposed to a bureaucratic process, focused on protecting organisations from prosecution. There should be no unnecessary documentation, no unread risk assessments and method statements, and no complex and often contradictory record-keeping systems.

He pointed out that the mechanisms of a minor accident, such as a twisted ankle, are not the same as a major incident, such as two cranes colliding. As such, focusing on minor accidents does not necessarily prevent the major ones. Safety professionals obsess about eliminating all accidents, being psychologically predisposed to avoid all harm, when really they should be focusing on the activities that can injure or kill.

It took Green several attempts to get the Australia Hub leadership to agree to trial the new approach, particularly as it meant moving away from the so-far highly successful zero vision. After much internal discussion, the company broached the subject with its workforce, the majority of which replied they did not believe a zero target could be achieved anyway – so no big deal. They accepted minor accidents would always happen in the ‘messy’ environment of construction, but that was OK as long as they got home in one piece every night. 

Encouraged by this, the company then asked employees what they liked and disliked about the new safety systems. They said they wanted to be more involved; given greater accountability.

Based on workforce responses, a series of moderated ‘micro-experiments’ were conducted – though the results were sometimes surprising. For example, on a brownfield site in Western Australia, with shallow foundations and no overhead working, the workforce was offered the opportunity not to wear hard hats, if that was the majority view. However, they declined, saying they would prefer to wear them rather than take a decision not to − evidence perhaps they had become accustomed to not thinking for themselves.

More successful was a micro-experiment to limit written assessments to high-risk activities only. This was tried on the Queensland Government’s New Generation Rollingstock project where the team took up the challenge of reducing paperwork, thereby eliminating most elements of the standard health and safety inspection programme. This enabled staff to concentrate on more impactful activities, such as workforce engagement and the management of major risks. The project was so successful that it became part of a government educational video, aimed at keeping young workers safe.

Revamped inductions were tried too. In place of the regular three-hour presentation, employees were given the option of a reduced version (covering basic statutory items) followed by a group discussion on the project risks and how they were going to deal with them. Most chose the latter, as it gave them a feeling of empowerment, allowing them to share their experiences and develop practical solutions. This exposed the management to a wealth of untapped expertise, resulting in better ways of working and better relationships.

Putting the theory into practice

The micro-experiments confirmed it: the new strategy had legs. It had the support of the workforce – and clients. In June 2015, the new strategy was formally rolled out across the Australia Hub. Each site was given a set of basic principles to follow, but allowed to decide the extent to which they wished to participate. This transferring of ownership to projects has so far proved successful, resulting in a groundswell of support from the workforce.

1: People are the solution, not the problem

The people responsible for putting processes into practice must not only have the competence, but the confidence to make the right decisions. To support this, the company has developed a tool called ‘collective insight’, which is used to guide group discussions between management and the workforce on high-risk activities. The aim of these sessions is to review the hazards involved and the associated control measures – and to invite feedback on the planned approach. Where viable improvements are proposed, they are incorporated into the methodology.

One idea initiated by the workforce, which has since been adopted, is ‘pit crews’ – committees of employee representatives, voted in by their peers, who meet regularly with management to discuss opportunities for improvement. The initial target is for there to be at least 20 per cent dialogue (as opposed to monologue) in all workforce encounters, but in time this should reach 100 per cent.

2: Safety is the presence of positives, not the absence of negatives 

There is now a strong focus on capturing examples of excellence so that these can be replicated across the organisation. To support this, the concept of a ‘positive investigation’ has been developed. The process applies standard causal analysis (used in the event of an incident) to activities that have been delivered safely. The aim is to identify the primary factors that contributed to its success, while isolating any redundant activities. For example, following the success of a major concrete pour on the Wheatstone LNG project in Western Australia, a positive investigation was conducted and a report produced. This was shared with the client, Chevron, who subsequently circulated it across its global operations.

Similarly, individuals can report positive observations through a ‘Good on Ya’ card. This initiative, which was instigated by the workforce, ensures those who go beyond expectations are recognised for their efforts. Other safety measures include traditional hard data, such as accident free hours and system health (evaluated in terms of overall performance and responsiveness to change) cultural data, such as frequency of feedback as well as from the workforce and degree of contribution in safety briefings.

3: Safety is an ethical responsibility, not a bureaucratic activity  

Safety management should be about protecting people from harm, not protecting companies from litigation. Paperwork and systems are enablers – helping us learn and share lessons. The gap between written guidance and actual work needs to be acknowledged – and workarounds and innovations explored.

More than six months into the new strategy, the DFIR in Australia is down to just 0.17, some 35 per cent better than the UK. On sites across the hub, one can see teams routinely coming together before and after the day’s work, planning their approach and (on successful completion) investigating the outcomes.

Commenting on its success, Dekker says: ‘It’s apparent that one of the few organisations that have effectively formed a practical view of the new safety paradigm is Laing O’Rourke. They have, in the words of John Green, ‘weaponised’ the theory and developed innovative and exciting tools to move forward’.

Bringing the change back home

Green is now back in post in the UK – and, after some convincing, the leadership has agreed to adopt the strategy in the Europe Hub. However, no one is under any illusion it will be easy, given the complexity of operations – not to mention the entrenched attitudes of the safety profession and regulators.

Certainly the UK-based clients Laing O’Rourke has spoken to so far have responded positively. They acknowledge the current approach is no longer working and are heartened by the prospect of an innovative yet practical alternative.

John GreenHealth, Safety and Environment Director

Download this article as a PDF by clicking on the image below.


Back to listing