One of the main hazards related to the exploitation of hydrocarbons is their high flammability and potential to generate fire and explosion scenarios. The calculation of the risks connected to such activities requires an accurate estimation of the ignition probability which is however an area of analysis still affected by a high degree of uncertainty. The most common approach makes use of UKOOA correlations which provide general curves of ignition probability as a function of release mass flowrate for different types of release environment. In November 2018, a new ignition probability model called MISOF was developed by Lloyd's Register specifically for offshore oil and gas installations located in the North Sea. Derived from a wide analysis of recent accidents, the MISOF model allows to capture and reflect the real time dependent evolution of the flammable cloud and its interaction with the ignition sources and ignition control measures in place (e.g. gas detection, isolation of ignition sources and depressurization of the process discharging equipment). The MISOF approach was applied considering four failure cases with different release rate of flammable material and hole sizes on a floating production unit (FPU). Ignition probabilities were estimated for the analyzed cases and compared with the figures obtained with the broadly used UKOOA correlations. The results showed that the ignition probability evaluation conducted with the MISOF model provides higher values than those obtained with the UKOOA correlations. By applying the MISOF, a significant increase in the frequency of final accident scenarios was found for Jet Fire and Flash Fire. Overall, the study demonstrated that the MISOF is a novel approach able to reflect in the analysis of fire scenarios the in-situ layout-specific knowledge of density and distribution of the ignition sources surrounding the release.
The evaluation of the safety status of an asset/production plant is mainly performed by the application of a Quantitative Risk Assessment (QRA) that, as a formal and structured methodology, aims at quantifying individual and societal risk related to all the potential hazards that could pose severe consequences to people. The framework of this analysis is carried out in a series of steps. First, all major hazards that are present at the specific site under investigation are identified; secondly the frequency and consequences that can arise from these hazards are separately evaluated; lastly, since risk is a combination of the frequency of the accidental scenario and the potential final consequences, these two elements are combined together into a risk index. Such a value shall be then compared to the risk tolerability criteria that are internally exemplified by the risk matrix in order to evaluate if the risk is acceptable or not. In case it is not, significant measures must be proposed in order to reduce it as low as reasonably practicable (ALARP). When performing the QRA analysis, there are many areas where the degree of uncertainty related to the quality of data and the way they are retrieved and further implemented may affect the overall calculated risk figures. In particular, the estimation of the ignition probability is a key point in evaluating the frequency of major scenarios involving flammable substances (i.e. Pool Fire, Jet Fire, Flash Fire, etc.).
