Journal of Loss Prevention in The Process Industries, Vol.9, No.5, 317-322, 1996
Application of ALARP to the design of the BP Andrew platform against smoke and gas ingress and gas explosion
In the UK, it is a legal requirement for a safety case to be produced for each offshore installation. A safety case involves a process called 'formal safety assessment' in which all hazards and risk to personnel on the platform are identified and quantified. The operator or designer of the installation is required to demonstrate that the risk to personnel is below a level which is deemed to be unacceptable, and in any case, reduced to a level which is as low as reasonably practicable (ALARP). in an existing facility, any risk reduction measures can be judged to be 'reasonably practicable' by balancing the cost of implementing the measure and the benefit in terms of risk reduction set against a baseline risk level (i.e. the current risk level of the facility). In the design of new facilities, there is no pre-existing baseline risk level. The demonstration of the application of ALARP is through the process of design. This paper describes the work undertaken during the design of the BP Andrew platform to assess and reduce the risk of temporary refuge (TR) impairment due to gas explosion and smoke and gas ingress. The process of platform design and risk analysis was integrated in the Andrew Project; safety technologists at BP using state-of-the-art quantification methods were part of the engineering design team at Brown & Root. This allowed the Project to identify practical changes in their design which could reduce risk significantly without incurring cost or schedule penalties. In fact, the project was able to identify design changes which reduce both the risk of TR impairment and cost of the facility, i.e. a negative cost measure! Two examples are included to demonstrate how the principle of ALARP was achieved by two different routes: (a) the platform was designed against a worst-case event, thus achieving the lowest possible risk, and (b) the risk of TR impairment of available practicable design options was accurately quantified and understood and the design option with the lowest risk was chosen. The project was able to demonstrate a final TR impairment of less than 4x10(-5) per year at the end of the design.