Human error is identified as a root cause of accidents and incidents in 60% to 80% of the cases and a causal factor in another 50% to 60% of all mishaps across many industries. Although the human contribution is recognized, efforts to improve human performance by influencing behaviors and attitudes are often ineffective or even elusive.
An indication of the continued need for effective development of safety cultures was the increase in fatalities reported in the “Safety Performance Indicators—2012 Data,” published by the International Association of Oil and Gas Producers (OGP 2013). OGP has been collecting safety incident data from its member companies globally since 1985, creating the largest safety database of safety performance in the exploration and production industry.
Based on data provided by 49 of the 63 OGP members covering operations in 107 countries, the fatal accident rate in 2012 increased by 27% compared with 2011, with 88 fatalities in 2012 and 65 in 2011 (Fig. 1). The fatalities, classified by activity, are shown in Table 1. Forty workforce fatalities were identified as being related to process safety events. The most common causal factors related to the fatal incidents and high potential events from 2010 to 2012 were
Fig. 2 represents the complexity of human factors that affect safety cultures (Poblete et al. 2014).
Although the classification and tallying of incidents and fatalities provides useful identification of where problems may lie, it does not address how to change the factors that led to them. Learning from incidents has been hindered by a tendency to treat human error as an acceptable final explanation of why an incident occurred. A deeper understanding of why human error occurred and the organizational/culture factors that set up the human for failure is required to make meaningful changes in safety performance.
The publication in March of SPE’s first technical report, “The Human Factor: Process Safety and Culture,” was intended to create a common understanding of the strategic challenges for the oil and gas industry, to identify what is known and unknown in the field, and to explore possible actions to accomplish the needed change indicated by the US National Commission on the Deepwater Horizon Oil Spill and Offshore Drilling Report.
Ken Arnold, senior technical adviser at WorleyParsons, introduced the report at a luncheon hosted by the SPE Gulf Coast Section’s Projects, Facilities, and Construction study group in January. He said, “Major accidents are not caused by failure of ‘safety critical’ equipment. We always need to be sure that our equipment and devices are safe, so almost everything is safety critical. If we are going to make a step change in safety, we have to look at something else.”
Arnold said the purpose of SPE’s technical report is to “present technical information on a topic where publication would be beneficial to the public and SPE membership. It evaluates available technology, but is not a recommended practice or standard.”
“Human factors” are defined as factors that can lead to the loss of standards of consistently reliable human performance that are relied on as part of an organization’s defenses against process safety incidents. The report notes that the definition should not be confused with the use of the term as a professional scientific discipline. As used in the report, the term refers to the factors that need to be managed and controlled to reduce the potential for major incidents, and not to the technical disciplines that bring specialist knowledge, skills, and techniques to mitigating these risks.
The report was based on the discussions and conclusions of about 70 participants representing a crosssection of the industry with individuals from oil and gas major operators, national oil companies, smaller operators, major contractors, regulators, universities, and consulting organizations after a 2-day summit.
The summit participants identified 10 human risk factors that the oil and gas industry needs to address. These are
Three issues were identified as the basis for progress in the others and as having the potential to be prioritized immediately within organizations: leadership and culture, perception of risk and decision making, and individual and team capacity. Dealing with these issues can go a long way toward achieving significant effectiveness across all aspects of a safety culture. Arnold said, “The challenge is to create and maintain an organizational culture that no longer allows things to happen when it is clear in hindsight that something should have been done and everyone knew it.”
The cultures of the involved organizations have been implicated in major process accidents, such as the Deepwater Horizon, a US Gulf of Mexico drilling rig (2010), where results of a negative pressure test were misinterpreted and the outflow from the well was overlooked; the Piper Alpha, a North Sea oil production platform (1988), where there was a failure to record the removal of a relief valve and a failure to recognize a gas alarm associated with putting a pump in service; and the P-36, an oil and gas production platform offshore Brazil (2001), where a valve was opened to a tank while a vent was blocked and watertight doors were left open after work was completed.
High-hazard activities were accepted as the norm, because no incidents had occurred in the past. In other cases, high-hazard well control events occurred with operating staff choosing to ignore, or accept as nonthreatening, signs of danger that they had been trained to observe. The acceptance of high-hazard conditions, even when people know things can be done better, is cultural.
It is often said that “safety begins at the top” of an organization. While valid, the statement should be qualified by the fact that there are multiple levels of decision makers between the top and the hazards, including the hands-on person(s) dealing directly with the hazard. Because the executives are often far removed from the hazards and the people dealing with them, their beliefs about the safety culture many not correlate with what is actually happening along the chain of command or on the ground.
Fig. 3 shows OGP’s health, safety, and environmental (HSE) culture ladder (2010), which represents the differences in organizational safety cultures. The introduction and use of HSE tools is affected by the type of safety culture in place, ranging from pathological to generative.
The SPE report notes that a crucial point is that people may want and support safety, but do not believe that it is fully achievable under profit, time, and production pressures. Arnold said, “While working on a USD 1 million/day drilling rig, seldom will a roustabout say that something is not safe. They don’t lend themselves to ‘stop work’ authority. Under time pressures, incomplete or wrong information may be relied upon to make decisions in situations where someone must act quickly, and applying actual knowledge in the real world is the sticking point.” The proactive and generative cultures (Fig. 3) are typically less vulnerable and more resilient, especially when processes do not proceed as expected.
Organizational leadership behavior must be assessed and assured. Training of leaders in the behaviors to assess during site visits has positive effects on improving process safety. Preparing leaders to identify what to look and ask for, and how to understand what the answers mean can provide them with a more accurate indication of the safety culture. By showing his or her ability to ask meaningful questions, the leader sends a message that safety is important.
The “Culture Matters” column in Oil and Gas Facilities regularly addresses the cognitive limits and biases in human decision making based on a range of scientific literature. The oil and gas industry has been inconsistent in implementation of the scientific knowledge about decision making and in many cases lacks awareness of the processes and limitations of cognition. Consequently, experienced, trained, and skilled people still make decisions that, in hindsight, appear to have been poor.
Arnold said, “We’re all trained in the mechanical, hard wire stuff, but not the soft wire.” The development of nontechnical skills is critical to improving safety culture. For example, situational awareness, conscientiousness, communication, cooperation and working with others, workload management, and self-management are important for managers to reinforce on an ongoing basis.
Decision making involves the whole process from awareness of a situation (perceiving a potential hazard or failing to do so), making projections of what might happen, planning possible courses of action, and making the right choice about what to do. Rational decision making models assume that people compare options and make optimized decisions when sufficient time and information are available. However, most of the time, decisions are made almost automatically, as when driving a car. Choosing something “good enough” for the moment and adapting one’s action, if necessary, is sufficient in most ordinary situations. But in safety-critical situations, it can be problematic or disastrous.
In addition to individual decision making, groups can also affect decisions. Group decision making is often slower than individual and requires energy and resources. It may not be clear who is responsible, leading to inaction or no one taking responsibility for the decision.
Riskier decisions may result when based on a group process. People who have a higher status can have a disproportionate influence on the final decision, causing those with more technical expertise to lack confidence in their opinions or decisions. If trained to question each other in a supportive manner, individuals in a group can counteract this tendency.
The situation in which a decision must be made cannot be dismissed. Conflicting goals, missing or unclear information, and time pressure can lead to poor decisions. In some cases, the complexity of operational processes may delay the feedback about the correctness of a decision. Safety-critical decisions may be subject to early finalization under time pressure or the perceived need to make a quick decision. Individuals, especially leaders, must be aware of the factors that increase the chances of making and accepting decisions that are risky.
A proper system design and/or timely and appropriate human action can usually be shown to have been able to prevent nearly all unsafe outcomes. However, the assumption that the ability to perform the correct actions is based only on adequate training and competency is false. Training, certification, and licensing are necessary, but not sufficient by themselves. Prevention of major accidents is dependent on adherence to the principles of safe design and appropriate safety management systems.
The SPE report notes that generally, there is a common understanding of the technical competencies required for key job roles in the industry. However, specific, documented, industrywide definitions of technical competencies do not exist. The definition and development of technical competencies will need to be continually reviewed and updated because of the increasing complexity of operations and technological advancements in the industry.
Nontechnical skills must not be overlooked. There is no industry agreement on the nontechnical skills needed for specific jobs. The current training available for skills in handling uncertainty and abnormal events must improve.
Some issues related to competency development can be addressed within individual companies, but because of the extensive use of contractor services and partnerships, an industrywide effort is necessary to identify the agreed upon skills, behaviors, and competencies.
For more information, visit www.onepetro.org to download the complete technical report, which is available at no cost to SPE members and nonmembers.
OTC 25130 Human Factors Analysis and Classification System (HFACS): Investigatory Tool for Human Factors in Offshore Operational Safety by J.D. Johnson and D.M.A. Hollaway, ABS Consulting.
OTC 25280 Human Factors in Hazard Analysis by B.R. Poblete, C.W. Parker, S. Ranasinghe, et al., Atkins.
International Association of Oil and Gas Producers. 2013. Report No. 2012s. OGP Data Series. Safety Performance Indicators–2012 Data, http://www.ogp.org.uk/pubs/2012s.pdf.
International Association of Oil and Gas Producers. 2010. Report No. 435. A Guide to Selecting Appropriate Tools to Improve HSE Culture, http://www.ogp.org.uk/pubs/435.pdf.
Society of Petroleum Engineers. 2014. SPE Technical Report. The Human Factor: Process Safety and Culture, https://www.onepetro.org/general/SPE-170575-TR.