Explosion Consequence Assessment
Background
Blast injury to people may comprise either direct effects (eg ear drum rupture) or indirect effects (injury due to flying debris). Blast damage to equipment of structures can result from either loading (applicable to large objects, eg walls) or drag loading (applicable to objects of narrow cross-section, eg pipework or primary steelwork) or a combination of the two. The extent of damage is dependent not only on the peak overpressure, but also the blast wave duration, impulse and rise time.
Of the 10 significant (> 0.2 bar) explosions which have occurred on offshore installations in the North Sea over the 25 year period 1973-97, all have resulted in significant damage to the installation, whilst 5 have caused injuries or fatalities (Vinnem, 1998).
Current position
Our understanding of the effects of explosions on people is, to a large extent, based on the observed far-field effects of conventional or nuclear weapons, which is of limited relevance to gas explosions on offshore installations. Our understanding of the effects of gas explosions on equipment and structures has advanced through the various phases of the JIP (Phase 2, Phase 3a and Phase 3b) each of which included some structural response experiments. However recent interpretation of some of this data has shown the importance of considering the response characteristics of a structure, as well as the explosion overpressure, in determining the loading imposed on such structures.
Modelling capabilities
Current models for assessment of blast injury to people are based on the far-field effects of condensed phase explosions (as noted above) and therefore their application to gas explosions on offshore installations is highly uncertain. The prediction of explosion loading on equipment and structures (and their response) is likewise highly uncertain. This has been shown through the various model evaluation exercises undertaken as part of the JIP. Other aspects of concern include the prediction of localised explosion loads, eg on objects of narrow cross-section such as primary steelwork, and the uncertain validation of blast wave codes. The modelling of escalation (eg through blast or missile effects) is also an area of significant uncertainty, with little detailed guidance available.
Industry practice
Safety Case assessments of blast injury to persons are largely judgmental, making use of what little data there is of relevance to the near-field effects of gas explosions. Assessments of the effect of blast on equipment and structures is sophisticated for new installations (eg using CFD for explosion prediction and non-linear finite element analysis for structural response) but highly variable for existing installations, ranging from purely qualitative analysis to the use of advanced computational models or experimental techniques. Escalation modelling is performed to widely-varying degrees of rigour.
Strategy development issues
Effects of blast on people
- Analysis of the effect of blast on people are highly uncertain as they are based on injury models developed from condensed phase explosions. They generally therefore to take a conservative approach, considering that all or a proportion of personnel in the vicinity of an explosion will be fatalities. A probit approach is not used due to the small distances involved.
- On offshore installations injury to personnel will be determined to a significant degree by secondary effects (eg collapsing structures, falling debris etc) and are not taken into account in risk assessments.
Effects of blast on equipment and structures
- The ability to predict, within reasonable accuracy, the explosion loads imposed on equipment and structures by hydrocarbon gas explosions is an essential element in the control of explosion risks on offshore installations.
- Currently there is uncertainty as to how to interpret overpressure data from large scale experiments for the purpose of understanding the loads imposed on equipment and structures, although it is anticipated that the Phase 3b work may shed further light on this.
- Current modelling capability falls short in a number of respects:
- the prediction of loads on objects which are small or of narrow cross-section (eg pipework, primary steel work etc) or the localised loads on larger objects; and
- uncertainty as to the validation of models used to predict 'far-field' explosion loads.
- Industry approaches to the assessment of explosion loads and structural response are of widely-varying rigour, particularly for existing installations.
- An initial study using coupled explosion prediction and structural response codes has shown that current (uncoupled) approaches may be giving significant overestimates of explosion loading.
Escalation modelling
- Escalation modelling is an important aspect of the hazard assessment and risk management of an installation especially in relation to TR impairment, giving key information on potential remedial measures to break a sequence of hazardous events.
- Treatments of escalation in QRA studies vary widely in their degree of rigour.
- Some potential mechanisms of escalation (eg missiles, partial blast wall failure, far field explosion effects on TR structures) appear not be given any specific consideration.
- Guidance on escalation modelling (UKOOA, 1995 and CMPT, 1999) is high level and broad in nature.