Flaws in the chemical sector’s approach to site safety have been dramatically exposed in a series of disasters in recent years.
An explosion at BP’s Grangemouth petrochemcial works in 2000 revealed the firm’s failure to detect and correct its deteriorating safety performance (ENDS Report 344, pp 6-7 ). A similar incident at Esso’s Longford gas plant in Australia in 1998 revealed shortcomings in the company’s safety monitoring procedures.
In response, the HSE is putting pressure on the chemical industry and other sectors with major accident hazards to improve their approach to managing safety. One of the main outcomes is a joint HSE-Chemical Industries Association guide on developing process safety indicators, aimed at senior managers.1
The initiative chimes with the government’s "revitalising health and safety" strategy. Launched in 2000, the ten-year plan aims to cut fatal accident and serious injury rates, which have not improved since the early 1990s.
The HSE’s guide says: "Too many organisations rely heavily on failure data to monitor performance… improvements are only made after something has gone wrong. Effective management of major hazards requires a proactive approach to risk management."
Companies typically focus on occupational injury data as the main measure of site safety. But the guide recommends measuring inputs as well as outputs: a dual system of "leading" as well as "lagging" indicators for each risk control system, such as staff competence, operating procedures, instrumentation and alarms.
For example, for inspection and maintenance, a lagging indicator would be the number of loss-of-containment incidents due to pipe, pump or storage tank failures. Corresponding leading indicators would be the percentage of safety-critical plant and equipment that performs to specification when inspected or tested, and the percentage of maintenance actions identified as being completed to a specified timetable.
Poor maintenance is a common cause of serious incidents. ConocoPhillips paid more than £1 million in fines and costs for an explosion at its Humber refinery in 2001. The incident was caused by a leak from a corroded pipe. An investigation revealed a "systematic failure" to inspect pipework in certain parts of the refinery (ENDS Report 366, p 56 ).
The HSE’s approach to developing process safety indicators was piloted with the help of BP, ExxonMobil and Syngenta.
Neil Macnaughton, process safety specialist at Ineos, which bought BP’s Grangemouth works in 2005, has been using the indicators for the past five years.
"There is no doubt that the indicators have driven improvements at the site," he said. "The HSE guide is world class."
He regards the way the indicators keep senior management informed about safety as one of the most important benefits. For each leading indicator, tolerances are set to trigger early intervention to halt deterioration in performance.
Other operators of sites with major hazards are under pressure to adopt the approach. HSE inspectors are routinely looking at whether companies are using process safety indicators.
The CIA sent the guide to its members and organised workshops to promote it. It also gathered information on the indicators currently used. The "vast majority" already had some process safety indicators, according to Colin Chambers, CIA head of Responsible Care, the industry’s safety, health and environment programme.
But many of the examples given are very general, and it is not clear whether they are expressed in a way that makes them useful as leading indicators. There is also uncertainty over whether company boards recognise the importance of monitoring key process safety indicators, rather than focusing on injury data.
The BP Texas City refinery disaster in 2005, which killed 15 people and injured 180, emphasised that lessons from Grangemouth have yet to percolate throughout the business. An investigation found that the company’s board had been led to believe all was well at the plant by indicators that were measuring the wrong factors (ENDS Report 382, p 10 ).
The prosecution of steelmaker Corus in December for safety failures leading to the 2001 Port Talbot explosion emphasises the need for companies in other sectors to act. The HSE found Corus had not responded to concerns over declining safety performance at the blast furnace, which eventually led to the incident (see p 51 ).