The results of extensive post-Buncefield research into the production of flammable vapour in incidents where large storage tanks of volatile liquids are overfilled were recently published in full.
The HSE-funded research report Flammable vapour cloud risks from tank overfilling incidents documents research undertaken in 2007 and 2008 but which was not published on completion due to constraints on Health and Safety Laboratory staff involved in the investigation arising from the legal case. The work was summarised for the Process Safety Leadership report but has not been published in full.
The huge blast at the Buncefield fuel depot in Hertfordshire on 11 December 2005 was caused by the overfilling of a large petrol-storage tank and subsequent formation of a flammable vapour cloud, which spread off site and ignited.
A series of investigations and reports were carried out on the incident; this latest report (RR937) includes much useful information arising from a joint study of tank designs carried out with Shell Global Solutions. The analysis describes how different tank designs are likely to behave in an overflow situation, and the impact on the production rate of flammable vapour.
Work in the area of Computational Fluid Dynamics (CFD) provides a first step towards developing a mathematical model to predict the size of flammable vapour clouds from overfilling releases, based on simple measures such as the tank dimensions, tank design type, pumping rates and liquid composition.
It also highlights the fact that there are important processes occurring at the bottom of the cascade near to the tank base that were not well understood at the time the report was originally completed, most notably splashing and pool evaporation (these issues have been explored in research report RR908).
A related report also recently published – RR936 Buncefield investigation: Liquid flow and vapour production – looks at the connection between the loss of containment and the formation of a flammable vapour cloud. Practical and numerical investigations have demonstrated that the bulk of fuel vaporisation and entrainment of air occurred during the cascading of fuel from the top of the tank into the bund.
The work involved the construction of a fullscale replica of a section of top of the tank involved at Buncefield and also a full-height section of the tank wall, to enable liquid flow experiments to be carried out.
How is a volatile fuel tank overfilled?? The fuel tank in my car cannot be overfilled owing to a back pressure arrangements that stops the pump when the tank is full. Are the operators saying that the Buncefield tanks did not have such an arrangement? Also did the driver of the tanker that overfilled the tank miss the available capacity indicators on the side of the tank, (assuming there were such things fitted!). This is shades of Camelford and the wrong connections.