TABLE OF CONTENTS Given the large number of possible fire scenarios for a given performance-based design project, it is usually necessary to reduce the possible fire scenario population to a manageable number of design fire scenarios for evaluating trial designs. Generally, possible fire scenarios can be filtered into design fire scenarios using the engineer's judgment on what fires will bound the potential hazards. If calculations are necessary, two general approaches for accomplishing this are available probabilistic and deterministic. Each design fire scenario (which is highly specific to support the hazard analysis calculation) is part of a scenario group (which is more general to support the frequency calculation) and is meant to be representative of that group. The scenario groups must collectively include all potential scenarios so that a valid risk measure can be calculated.
A probabilistic approach typically deals with the statistical likelihood that a fire will occur and the outcome if a fire does occur. The decision whether to select a given possible fire scenario as a design fire scenario is based on grouping scenarios that are similar. Grouped sets of scenarios can potentially be eliminated from further evaluation if the stakeholders agree that the risk or outcome is acceptable. A probabilistic approach could use the following as sources of data:
Fire statistics include statistics that identify the most likely areas of ignition, items first ignited, and the likelihood of spread beyond the room of fire origin.
