Despite the general perception of 2020 as an “annus horribilis”, there were also some optimistic safety trends with injury frequency relatively low. Based on industry statistics, several companies are expected to set an all-time low record in Lost Time Injury Frequency (LTIF) and Total Recordable Injury Frequency (TRIF). Even so there are still too many high potential incidents*. Some say that the underlying causes of several of these high potential incidents were cost-cutting and delayed maintenance, enhanced by low oil prices and COVID-19. But there are alternative explanations. A biased focus on minor incidents and rigid control can easily be prioritised at the expense of preventing far worse major incidents. Without insight and due care, it is easy to go down the wrong path. This article outlines some of the challenges and possibilities for guidance in this difficult terrain.
Norwegian PSA investigation findings
The Norwegian Petroleum Safety Authority (PSA) is tasked with investigating potential and actual major incidents. Its investigations into the recent 2020 incidents are not yet completed. However, a review of 68 investigations identifying more than 140 failures (deviations from what is expected, desired, or required) demonstrates some clear patterns.
The failures which involve how the personnel actually do their work (red) are far more frequent than the structural “non”-human systems they have in place to prevent major incidents (blue).
The largest single failure identified is “Risk management in planning and execution”, which was found in more than a quarter of the incidents. In contrast, maintenance has played a key role in only nine of these investigations. Although the recent high potential incidents may involve more maintenance issues, the major takeaway from these investigations is that behavioural failures have greater improvement potential.
Injuries and major incidents have different patterns
The relationship between injuries (High-Frequency, Low-Consequence (HFLC) and major incidents (Low-Frequency, High-Consequence (LFHC) involves several misconceptions and conflicting interests. The investigations into two major incidents, the Texas City explosion in 2005 and Macondo in 2010, concluded that the management had a biased focus on injuries such as LTIF at the expense of major accident risks. Injuries and other HFLC incidents are usually triggered by single failures and the injured person is normally the one who makes the mistake. Innocent parties are seldom exposed.
Major incidents, on the other hand, have different characteristics involving multiple failures, several of which are caused by persons in different departments and at various levels in the organisation. The failures are often unaddressed over time until an unfavourable situation results in major losses that frequently involve innocent persons.
These incidents are therefore often referred to as organisational accidents. Because of the difference in the nature of injuries and major incidents, there is a need to be more aware of, and up to date on, the various strategies to prevent these incidents from happening.
Different safety strategies have different effects
In other industries, a biased focus on structural interventions and transactional implementation results in a doubling of the frequency of major incidents while the injury frequency is halved. This emphasises the need to more thoroughly evaluate how injuries and major incidents can effectively be prevented.
The first distinction should be between structural and behavioural interventions. Structural interventions, such as requirements for the use of personal protective equipment (PPE), can be used to prevent injuries. Similarly, barrier management system is a structural approach to prevent LFHC incidents.
The second distinction is between various ways of implementing changes. There are several different ways, but for simplicity let us consider the two most relevant, which are the transactional and transformational leadership approaches.
A transactional implementation tends to focus on supervision, responsibilities, rewards, sanctions and authority. The underlying driver is to “control” the implementation process. In effect, it uses power to change behaviour, often under the mantra: what is measured gets done. An alternative to a transactional approach is the transformational leadership approach.
Transformational leaders aspire to the “change” itself over “control”. Transformational leadership wants to build ownership of the change, motivate for goal achievement, inspire innovation, and show concern for the employees’ welfare. Several studies demonstrate that a transformational leadership approach is more effective in achieving sustainable behavioral change.
Some key ingredients of safety improvements
So, back to the distinction between injuries and major incidents. Because of the characteristics of individual injury accidents (HFLC) described above, it is obvious that a transactional approach may have some effect on compliance with procedures and other requirements. For example, if you require people to use personal protective equipment (PPE), the injury frequency will decline. If you supervise everybody and sanction the ones that do not use proper PPE, the HFLC statistics will probably continue to improve.
In the energy industry, which has potentially high operational risks and decades of experience, there is a foundation of accepted construction and regulation in place. Hence, as the PSA investigations reveal, the need for additional “structure” is not that evident in the prevention of major incidents (LFHC).
A socially safe working environment based on trust, care, and openness is required to allow people to share their own mistakes and address failures caused by their colleagues or even managers. If people believe they will be fired for speaking up, they will keep their mouths shut.
Trust, care, and openness cannot simply be required (like the wearing of PPE). On the contrary, this kind of environment is something that managers and leaders must create. Hence, there is an evident need for behavioural interventions combined with a transformational leadership approach.
Sometimes a quick fix is tempting
With a strong public focus on sustainability and transparency, all energy companies are committed to continuously improve their safety performance. Injuries and other high-frequency low-consequence incidents can be a substantial problem and are easy to monitor due because they occur more often than major incidents. Companies that fall behind on improvements must expect negative attention from investors, media and the public.
One tempting “quick fix” to improve the injury frequency is to implement structural interventions through a transactional approach; write a new rule and simply require people to comply. Because this approach has long traditions, it is referred to as the “old school” approach. A transformational approach requires more cross-organisational collaboration between the safety department (which “owns” the problem), the HR department (which “owns” the people) and day-to-day operations (which “own” the managers and priorities). Furthermore, their often involves more advanced implementation. Bottom-up engagement ensures that the employees have ownership and understanding of the change. The top-down implementation through the “chain of command” ensures that the changes are welcomed.
Old school safety management fuels cover-ups
Many energy companies unconsciously rely too much on structural interventions and/or transactional implementation. For example, to speed up reporting it is common to require a certain number of observations or near-miss reports per employee or installation per month. Other examples are behaviour-based safety and barrier management, which are often implemented using a transactional approach.
When people are forced into a change without buy-in, engagement, and motivation, the change can easily turn into a compliance exercise without ownership and critical thinking. The end result is that complying with the requirement or expectation tends to be more important than the underlying purpose and intention.
The reporting of lost time injuries is one example of under-reporting due to both the internal and external transparency regarding these incidents. Some typical strategies that stimulate the under-reporting of lost time injuries are listed below:
- Bend the definition: creatively find any simple task that makes it possible for the injured person to continue working.
- Suppress reporting: suppress the reporting of injuries by using indicators such as “days since last LTI” or “7 years without reported LTI”.
- Swapping: define the location of the injury to be outside the workplace or outside working hours or change the project that the injured person is working on.
- Simple under-reporting: simply violate the requirement to report an injury unless it is impossible to conceal.
Energy companies of the future will embrace transformational leadership
With few exceptions, the failures leading to major incidents are known before the incident takes place. This underlines the need to build a culture in which people dare to speak up and where they listen to each other. Transformational implementation is effective in maturing trust, care, and openness in the organisation, resulting in a 50-80% drop in major incident risk.
These are significant improvements, but there are other benefits as well. A more mature culture also has an impact on the serious injury frequency, which is reduced by a factor of 40-65%. Employees who have the buy-in and understanding take greater care of themselves and their colleagues.
So, when somebody asks if they should focus on major incidents or personal injuries, it is simply the wrong question. The most effective approach is to address both of them with a common transformational safety program.
* Examples of 2020 incidents are fires and explosions (Marathon, ExxonMobil, Equinor, Astron and Engen), blowouts (Assam/Baghjan (India) and Orenburg (Russia)) LNG process plant problems (Equinor and Petronas) and rig problems due to heavy weather (Seadrill/Lundin and Petronas Carigali).