Modeling and Assessing a Spectrum of Accidental Fires and Risks in a LNG Facility

Overview

Main Objective

The objective of the research proposed is to review potential LNG release scenarios in storage terminals and from LNG ships, and augment exiting analytical/computer models or develop new ones for different types and sizes (a spectrum) of LNG fires by properly considering important phenomena that have effects on the fire characteristics and the hazards they pose. A second objective is to develop protocols for using these models in performing a risk assessment of LNG transport in ships or storage in terminals. Additional objective is to provide mathematical tools with which to make Regulatory assessments or LNG terminal siting decisions.

Public Abstract

Abstract: Several applications for constructing and operating US based LNG import & storage terminals are being evaluated by US regulatory agencies. The assessment of risks to the public from these proposed facilities requires the consideration of thermal exposure hazards from a spectrum of potential LNG release fires arising from releases of LNG either by accidents or by terrorist actions. This proposal addresses the development or improvement of LNG fire models for different types of fires and coding the models for execution on a PC level computer. Current models for large LNG fires do not consider the effects of several important phenomena such as fractional distillation of LNG, black soot formation and consequent (fire shrouding) reduction in thermal radiation, decrease in visible flame length, etc. These phenomena will tend to reduce the hazard distance. Use of existing models for postulated large LNG fires has resulted in the prediction of very large hazard areas leading to safety concerns expressed by public interest groups and regulatory agencies. Developing models, using available field test data, and including these real phenomena that occur in large, turbulent LNG fires is also proposed. Regulatory agencies and the LNG industry can utilize the models immediately for decision-making.

Summary and Conclusions

A semi empirical, large LNG pool fire model, termed the Pool Fire Model Including Smoke Effects ("PoFMISE"), has been developed to predict the thermal radiation output from large hydrocarbon liquid fuel pool fires. This model takes into account the production rate of smoke (and the smoke's shrouding effect on thermal output), the variation of emissive power from the bottom to the top of the "visible" fire, and the presence of and variation with size of the bottom "bright burning" zone. Using the model, the fire-geometry-based mean surface emissive power (MSEP) for LNG fires of various diameters can be calculated.

Additionally, a methodology of evaluating the risks from potentially large LNG pool fires is illustrated. This methodology describes the various phases in the performance of a risk assessment, the identification of fire scenarios, and the development of the various probabilities of occurrence of the scenarios. Also illustrated is the calculation of the consequence areas and people exposure to hazardous radiant heat flux after considering the mitigating and shadow effects of buildings and structures, and the atmospheric absorption of radiant heat.

Relevant Files & Links