HABER

from safety culture to culture for safety

What is it that we still have not learned?

S. B. Haber

Human Performance Analysis, Corp.

New York, NY USA

Email:

Abstract

In April 1986 the Chernobyl Accident happened. In 1991 the IAEA Independent Nuclear Safety Advisory Group published INSAG 4 and the concept of safety culture was defined for the nuclear community because of its relationship to the accident. Work in the nuclear community evolved around the concept of safety culture although a clear understanding of what was actually meant was often missing. Methods to evaluate and assess safety culture were developed. Safety culture became thought of as a process that could be written into a procedure, measured by performance indicators and fixed in a corrective action program. Short term improvements in safety performance and the metrics to measure them were observed and many concluded they had really changed their safety culture. The changes were often not sustainable. The efforts did not include an in depth understanding of why individuals thought or behaved in the way that they did. In March 2011 the Fukushima Daiichi Accident happened. The IAEA conducted an in-depth analysis of the human and organizational factors of that accident and drew a number of conclusions but none as critical as the finding that while the same natural phenomena might not occur in every nuclear facility around the world, the same human and organizational issues could. What is taken for granted and what is assumed represents culture and will influence behaviors, decisions, and what is attended to. By working within the organizational culture to achieve and maintain safety performance a more realistic and sustainable outcome will result. It is time to shift the thinking in the nuclear community from safety culture to culture for safety. It is the necessary step for each organization to try to move forward to achieve long term sustainable safety performance.

1.  INTRODUCTION

In April 1986 the Chernobyl Accident happened. Several years later in 1991 the IAEA Independent Nuclear Safety Advisory Group published INSAG 4 [1] and the concept of safety culture was defined for the nuclear community because of its relationship to the accident. Where the Three Mile Island Accident in 1979 had brought human factors issues in procedure development, human performance, and training to light, the Chernobyl Accident was discussed in terms of management, supervision, and safety culture. Work in the nuclear community evolved around the concept of safety culture although a clear understanding of what was actually meant was often missing. Methods to evaluate and assess safety culture were developed and efforts to integrate the findings of those evaluations into more traditional nuclear tools, such as probabilistic risk/safety assessment were attempted as well. Safety culture became thought of as a process that could be written into a procedure, measured by performance indicators and fixed in a corrective action program. The changes that organizations saw as a function of their safety culture improvement programs though were often just changes in some behaviours. Short term improvements in safety performance and the metrics to measure them were observed and many concluded they had really changed their safety culture. The changes were often not sustainable. The efforts did not include an in depth understanding of why individuals thought or behaved in the way that they did.

In March 2011 the Fukushima Daiichi Accident happened. Initially it was accepted to explain it as a natural disaster. While the earthquake or the tsunami could not be prevented, there were things that could have been done before, during and immediately after the natural phenomena that would have helped to mitigate the consequences of the accident. The IAEA conducted an in-depth analysis of the human and organizational factors of that accident [2] and drew a number of conclusions but none as critical as the finding that while the same natural phenomena might not occur in every nuclear facility location around the world, the same human and organizational issues could. What is taken for granted and what is assumed represents culture and will influence behaviours, decisions, and what is attended to. There is a need for the nuclear community to move forward now in the way it thinks about safety culture. By thinking about culture as the foundation for the shared beliefs and values in any organization the realization that improvement programs will not succeed with short term efforts but rather will require time and commitment, will be evident. By working within the organizational culture to achieve and maintain safety performance a more realistic and sustainable outcome will result.

2. THE PAST 30 YEARS

After the accident at Chernobyl many efforts especially across North America and Western Europe were initiated to better understand the influence of organization and management on safety performance. The challenge was in recognizing that this was an area that had not been relevant for the nuclear community up until this point. Experts in fields outside of the traditional areas associated with nuclear technology had not been involved in ensuring nuclear safety.

From the North American perspective, researchers at the University of California at Berkeley, were involved in studying ‘high reliability organizations’ [3]. High reliability organizations were those that could not afford to make a mistake because the consequence to public health and safety would be significant. Initially their work focussed on naval aircraft carriers and air traffic controllers. Concurrently, the U.S. Nuclear Regulatory Commission (NRC) began research into the influence of organization and management influences on performance reliability in nuclear power plants [4]. In the late 1980’s the two sets of researchers worked together and conducted a study that involved both a fossil and nuclear power plant for the feasibility of a methodology that could evaluate these organization and management factors. For the NRC researchers the goal was to develop a methodology that would allow the quantification of these factors so they could be incorporated into probabilistic risk assessment. For the Berkeley researchers the goal was to continue to delineate the characteristics of high reliability organizations. Both objectives would eventually become critical in the evolution and implementation of the concept of safety culture.

Over the next several years the NRC researchers continued their work and published the results of their efforts [5]. A systematic, more objective and standardized methodology for assessing the influence of organization and management on safety performance was developed and implemented. When the NRC decided to discontinue the work, their own Advisory Committee for Reactor Safeguards published comments that recognized “management science as a real and sophisticated academic field that needs to be tapped if the industry is to continue to make progress in dealing with organizational performance issues.” The comments go on to say that “There appears to be a lack of communication between the management science academic community and most policy-makers out in the ‘real world’ of nuclear power plant regulation and operations. We believe that the Commission should encourage the involvement of the management science community in helping to improve the organizational performance of both the staff and the nuclear utilities. [6]” The NRC discontinued the management science aspects of the project for several reasons and the U.S. commercial nuclear industry took responsibility for ensuring that the management aspects of the nuclear plants’ performance would promote and ensure safety [7]. Efforts were being made in other countries to continue such work and use the methodology developed by the NRC research to promote their own objectives [8]. In addition, a review of the theory and research around the nature of safety culture was published in 2000 [9].

In 2002 the Davis Besse Reactor Vessel Head event in the U.S.occurred. In spite of issues identified by employees over years and bulletins about the existence of similar problems across the industry, management at the First Energy Corporation allowed the problem to continue. During the 2002 refuelling outage a ‘hole’ in the reactor vessel head created by boron leakage had gotten to within ¼” of the lining of the vessel. The causal analysis of the event identified ‘safety culture’ as the root cause. It was the first time in the U.S. commercial industry that the term safety culture was used as the basis for a failure in the performance reliability of a nuclear power plant.

Around the same time as the Davis-Besse event, the International Atomic Energy Agency (IAEA) published a document identifying the characteristics in an organization necessary for a strong safety culture [10]. The IAEA described 5 characteristics along with multiple attributes for each characteristic. In 2004, two years after Davis Besse, the Institute for Nuclear Power Operations (INPO) in the U.S. published, ‘Principles for a Strong Nuclear Safety Culture’ [11]. INPO identified 8 principles of a strong nuclear safety culture. The U.S. NRC published ‘Safety Culture Components and Attributes’ in 2006. The NRC identified 13 components with multiple attributes associated with each component. In 2014, 28 years after Chernobyl, and 12 years after Davis Besse, the U.S NRC, INPO and other nuclear license holders in the U.S. agreed upon the ‘Traits for a Healthy Safety Culture’ [12]. Ten traits were identified with multiple attributes for each trait. In 2015 the IAEA initiated an effort to get the IAEA, INPO/NRC, and the World Association of Nuclear Operators (WANO) to agree upon a common language with respect to safety culture; that effort is ongoing.

During the time of the development of these various frameworks for safety culture one thing was clear; most organizations did not really understand the concept, know how to assess it or what to do about it if change was required. The concept of safety culture became treated as a discrete entity that could be easily measured, managed through performance indicators, and ‘fixed’ with a corrective action. Safety Culture Improvement Programs often consist of hundreds of corrective actions that organizations believe can be completed individually and independently to obtain the desired results. Experience has shown that what these actions and programs are really changing is safety performance as represented by expected behaviours and outcomes. Often these changes are only short-term because the basis of the behaviour, and consequently the performance, is not understood well enough to ensure the changes are long term and sustainable.

3. EMPHASIS SHIFT – WHY CULTURE FOR SAFETY?

The term safety culture should implicitly create the necessity to identify and understand the concept of culture as much as the notion of safety. This has been a weakness in the thinking to date. Schein published a model of organizational culture which has become widely accepted and describes three levels that comprise culture; artefacts, espoused values, and basic assumptions [13]. The artefacts are the observable outcomes that are representative of culture that can be seen, e.g. behaviours, practices, and policies. The espoused values are the claimed shared values and norms that the organization communicates. The basic assumptions reside in the lowest level of culture and are the most difficult to capture. As Schein defines culture: “The shared basic assumptions that have come to be taken for granted and that determine the member’s daily behaviour.” [14] Basic assumptions are often unconscious or subconscious and are reflected in behaviours that are done automatically without reflection or analysis. It is from this level that culture drives organizational behaviour which results in performance and impacts all aspects such as safety, security, production or economics.

The idea of basic assumptions is best explained through an example. Assume that an organization shares a basic assumption that its employees are productive, creative, and motivated. When someone sees an employee sitting at their desk and staring out the window the basic assumption will drive the belief that this individual is thinking about a new idea or trying to resolve a problem. If the same organization, however, shares a basic assumption that its employees are lazy, not creative and non-motivated, then when someone sees an employee sitting at their desk and staring out the window the assumption will drive the belief that this individual is daydreaming, wasting time and being non-productive. Same behaviour, sitting at the desk and staring out the window, very different conclusions. The importance of basic assumptions in understanding behaviour and the reasons that certain actions occur cannot be over stated.

By better understanding the concept of culture it becomes clear that the idea of ‘safety culture’ and how it has been conceived, assessed and managed to date, is not getting at the lowest level of culture, the basic assumptions. It is the artefacts and some of the espoused values that have been assessed, measured and managed by most organizations. By not getting at what drives the behaviour that is observed, the changes that are made become short lived and often revert back to their original form not long after corrective actions have been implemented. This creates the cyclical nature that is often observed in performance in many organizations. The intention of organizations in trying to implement safety culture improvement programs is by and large sincere, but without the incorporation of an understanding of the organization’s culture and how it drives performance, the results are not likely to be effective over the long term.

The idea of ‘culture for safety’ more accurately reflects what must be understood and described for an effective and sustainable change in safety performance. Thinking back to some of the original definitions of ‘safety culture’ [9] we can create a working definition of ‘culture for safety’. Culture for Safety refers to “the characteristics of the work environment, such as the values, rules and common understandings that influence employees’ perceptions and attitudes about the importance that the organization places on safety” [15].

3.1 Methodological Premises

To describe a Culture for Safety accurately, an organizational cultural assessment

must be conducted, whose focus is safety performance. This implies some methodological premises:

-  Models of organizational culture identify behaviours as the observables of the values and beliefs underlying them.

-  Understanding the behaviours and having reliable and valid tools to assess and describe them is the most effective way to understand and assess organizational culture.