Benchmarking Safety Culture in Major Hazards Industries in the Rotterdam Area

Benchmarking safety culture in major hazards industries in the Rotterdam area (Netherlands)

By Gerard I.J.M. Zwetsloot[1], Robert A. Bezemer[2], TNO.

Introduction

Safety culture is increasingly recognized as an important determinant of safety, complementary to technology and safety organization. Safety culture is in itself a 'soft' notion that may have, however, ‘hard' safety implications. The relevance of safety culture was firstly addressed in the aftermath of the Chernobyl disaster (IAEA 1986) and the Piper Alpha disaster in the North Sea (Cullen 1990). We define safety culture as the attitudes, values, (implicit) assumptions, perceptions and practices with regard to dealing with safety risks.

With respect to the safety organization and technology, major hazard companies have to comply with rather detailed legislation. In the European Union this concerns the so called ‘Post Seveso (III)' legislation. There are no legal requirements with respect to safety culture, except for nuclear power plants. Safety culture is and remains therefore a responsibility of the companies involved.

Several incidents in the Netherlands in companies dealing with large quantities of chemicals got a lot of attention. It was broadly recognized that the safety cultures in these cases were relatively poorly developed. As a result there is a growing societal interest in the Netherlands for early recognition of serious safety risks and poor safety cultures in order to prevent future similar incidents. There is also ‘societal pressure' on chemical industries and government to take adequate preventive action.

Against this background, the regional environmental inspectorate DCMR asked TNO in 2012 to assess the safety culture in fourteen major hazard companies in four related sectors. The focus was on those characteristics of safety culture that are relevant for environmental safety and the management of major hazards.  

Research objectives

The two main objectives were:

•    To gain insight into the safety culture of 14 companies in four industrial sectors: refineries, (petro) chemical industry, bulk storage (tank parks), and chemical warehousing and logistics;

·      To allow benchmarking of safety culture between those sectors and between companies

Methodology

The "TNO Quick Scan Safety Culture" was conducted in fourteen companies. The TNO Quick Scan measures 14 dimensions of safety culture; nine dimensions are taken from the Hearts & Minds methodology (which comprises 18 dimensions, Parker et al 2006) as that methodology is broadly used by the chemical industry. Five other dimensions are included because the scientific literature and industrial experience demonstrate that they are very relevant for process safety.

The Fourteen dimensions of process safety culture measured with TNO's Quick Scan Safety Culture
Dimensions from the Hearts & Minds methodology (Parker et al 2006)	Complementary dimensions
·         Leadership and commitment ·         Safety communication ·         Opinion of management about the causes of incidents ·         Profit versus safety ·         Participation and commitment of employees ·         Contractor management   ·         Procedures ·         Incident reporting and analysis ·         Execution and follow-up of audits	·         Learning from incidents ·         Personal versus process safety ·         Handling complexity ·         Maintenance management ·         Functioning and roles of supervisors

The method comprises the following three elements:

- Preparatory stage: assessment of historic data about safety incidents, safety programs, and correspondence with the local environmental safety authorities (including non-compliances and fines), based on information from the inspectorate and the company.

- Two day assessment at the site, involving two researchers. The assessment comprised a walk-through (observations) and a series of interviews, at least with the plant manager, HSE manager, middle managers and supervisors, operators from day and continuous shifts, personnel of contractors, maintenance manager or personnel, and a member of the Works Council. In each interview all dimensions of safety culture relevant for the interviewee were addressed; this implies that with e.g. the plant manager and the HSE manager all 14 dimensions were addressed, while in other interviews only a selection of the dimensions was addressed.

The outcomes of the TNO Quick Scan Process safety culture (per company)

A better understanding of their safety culture, in 14 dimensions

A score on the safety culture ladder (similar to that in the Hearts & Minds methodology ), a scale from 1 (pathologic safety culture) till 5 (generative safety culture)

Identification of strengths and weaknesses of the safety culture

An assessment of the consistency in the safety culture: variation in quality of safety culture among the 14 dimensions and the degree of consistency between the various interviews

Often: useful observations about the functioning of the Hazard and Risk Control System in practice

A substantive dialogue between management and researchers about the findings and their meaning for the company's process safety management

- Analysis and reporting stage: the results of the various companies were analyzed, benchmarked and integrated into one report.

A team of seven researchers, consisting of experienced safety and psychosocial experts having a broad experience with safety culture carried out the assessments. They were trained in advance to be prepared to deal adequately with socially desirable responses of the interviewees, and to make sure they would get insight into the real safety practices.

Checks and balances between different sources (triangulation between observations, documents, interviews and between interviewers) is an essential part of the methodology. For each company the TNO Quick Scan Safety Culture was concluded by a final meeting in which the results were presented to management and were discussed.

Findings

Scores

The safety culture scores, which are the most important for benchmarking, show that the (petro) chemical companies and the refineries all have a good or acceptable safety culture. Whereas a score of 3.0 (a calculative safety culture) is regarded as the minimum acceptable score for a major hazard company, several companies in bulk storage and chemical warehousing/logistics score this minimum or even lower.

It is also possible to identify which dimensions of process safety culture were well or poorly developed in each company and in the four industrial sectors. Obviously, in the sectors with a weaker safety culture, more dimensions of safety culture were weak. In the (petro) chemical companies and the two refineries no weak dimensions were found. However, for the bulk storage and the chemical warehousing several weak dimensions were identified, while excellently-developed dimensions in these sectors were scarce.

Scores on culture ladder per sector
Sector	Average Culture score per company	Range
Refineries (N=2)	3.7	3.4 – 4.0
(Petro)Chemical (N=4)	3.9	3.5 – 4.0
Bulk storage (tank parks) (N = 4)	3.3	3.0 - 4.0
Chemical warehousing and logistics (N=4)	2.8	2.5 – 3.0

Consistency

In five of the 14 companies the score on each of the 14 dimensions as well as all interviews indicated a good safety culture. The average culture score of these five companies was 3.8. These companies can be regarded as frontrunners in the development of safety culture.

Three other companies also achieved an average 3.8 score, but in these organizations there were remarkable inconsistencies between the various interviews, i.e. there seemed to be subcultures or ambiguities in the safety culture. Unfortunately, within the limited timeframe available for the Quick Scan, it was not possible to further investigate these inconsistencies; as a result the good score of these three companies is associated with greater uncertainties.

Of the remaining six companies, one had a very consistent score of 3.0. Three other companies also got an average score of 3.0, but in these cases the consistency among the dimensions and interviews was less. Finally, there were two companies with a score of no more than 2.5. In these companies there was a range of weakly developed dimensions of safety culture; there were complementary observations of some specific unsafe practices as well.

In a few companies there were observations of some specific unsafe practices. Again, we were not able to further investigate this, which means that for these cases the safety culture scores are associated with significant uncertainties.

In each company and each sector there were opportunities for improving the safety culture. Three dimensions could be improved in all four sectors: learning from incidents, dealing with audits and reviews, and dealing with complexities.

Additionally, in the tank parks and the chemical warehousing the management vision on accident causation as well the reporting and analyses of incidents can also be improved.

In the chemical warehousing and logistics, three other dimensions also require greater attention: safety leadership and commitment, safety communication, and participation and commitment of employees.

Discussion

Limitations of this research

The method used in this research was the TNO Quick Scan Safety Culture. There are clearly limitations due to time limitations of this method. Nevertheless, the method proved to be useful for benchmarking purposes and for providing the companies with useful feedback on the strengths and weaknesses of their safety culture.

Opportunities for improvement

From several companies we got the feedback that the benchmark provided them with valuable insights into specific issues in their safety culture, which was used by management in their plans to (further) improve their safety culture.

---

[1] Gerard Zwetsloot PhD, is senior research scientist at the Netherlands Organization for Applied Scientific Research TNO. He is also honorary professor at the University of Nottingham (UK, China, Malaysia) were he holds a chair in Occupational Health and Safety Management.

Email: gerard.zwetsloot@tno.nl

[2] Robert Bezemer MSc MTD, is research scientist at TNO

---