Two years ago, the US refining industry suffered its worst accident in 20 years at Texas City, causing 15 fatalities and injuring 170 persons. In relation to this, two major investigations recently produced a number of valuable recommendations to the refining industry. However, DNV is not fully convinced and suggests all oil and gas industries should go beyond these reports in order to achieve a major improvement in business and safety performance.

The Baker Report in January 2007 looked at the safety management systems and culture inside BP’s US refineries, and in March the US Chemical Safety Board reported on the accident itself. Both assessments generated numerous valuable recommendations.

The Texas City accident should act as a wake-up call to the refining industry, as it demonstrates that major assets are insufficiently protected by current safety processes and investment strategies. The good news is that trends in personal injuries in the refining industry have been very good, with a major improvement since the early 1990s. However, the trend in major refinery accident damage has been poor to mixed, with many serious incidents and increased losses over the past five years.

Very few of the root causes of the Texas City accident were unique, and many have been seen several times before in major accidents. Also, several other incidents with lesser impacts could have escalated to the same severity in other locations had the circumstances only been a little different.
So not only the downstream and process industries can learn key lessons from this accident; other industries should be looking at the opportunities stated in the reports to improve performance and safety. For the upstream industry, for example, there is much to learn from a careful review of these two reports, and a more general examination of whether its personnel and assets are sufficiently protected against large process accidents.

Go beyond recommendations
Despite the two reports, DNV believes that these did not fully recognise hard-won lessons learned from other accidents and thus are not, on their own, the holistic approach necessary to achieve the next major step-change improvement in process safety.

In addition to implementing a broad risk-based programme, DNV suggest companies to focus on:

• Leadership: Should increase their top management’s commitment to, and awareness and leadership of, process safety to match that achieved today for personnel safety.
• Risk management: Should have more focus on process safety issues through tools such as risk assessment and risk registers.
• Barriers and controls: Should systematically identify key barriers and controls – both hardware and administrative systems and establish control accountability.
• Processes: Should have in place evergreen management processes to control both hardware and procedures.
• Leading indicators: Should have suitable and easily understood process safety metrics (leading indicators) that are regularly reviewed by management in order to monitor and benchmark risks.
• Verify effectiveness: Should initiate a fuller use of accident investigation to verify a control’s effectiveness.
• Closing: Should establish solid action prioritisation, follow-up and close-out tracking systems.

With this in place, gains of 2–3 times the current process safety performance should be achievable, so the measures are very cost effective. By increasing reliability and uptime, the assets also deliver improved business performance.

It is fortunate that the need for such improvement arrives at a time when industry has adequate resources due to the high oil price, thus having the opportunity to secure the assets and invest for the future.

The lessons are clear, and there is no longer any reason for the industry to delay action.