Comparative Safety Assessment of Prescriptive Means-of-Egress Configurations Using Fault Tree Analysis

Abstract Prescriptive building regulations commonly permit several means-of-egress and rescue configurations that satisfy the same regulatory safety objectives without explicitly quantifying whether these configurations provide comparable safety. This study develops a comparative fault tree analysis (FTA) framework and applies it to 21 residential means-of-egress configurations derived systematically from the German Model Building Code and related provisions. A common top event is defined as an acute fire service rescue being required before fire extinguishment outside the fire dwelling. The models contain 93 basic-event input values. All 93 inputs are assigned ex ante by expert judgement using a seven-level, ordinally anchored numerical likelihood scale, with five deliberately selected intermediate values where finer differentiation was required. They are therefore interpreted as comparative numerical likelihood inputs rather than empirically validated failure probabilities. The objective of the study is not to establish consensus probabilities but to investigate whether differences emerge under a consistently applied expert-based assessment framework. The resulting top-event metric is consequently reported as an FTA-derived score. Across all 21 configurations, the score ranges from 0.274 to 0.852. More importantly, among the 14 standard-code configurations that are treated as regulatory alternatives, the range is 0.477 to 0.809, corresponding to a comparative score ratio of 1.70. A monotonic scale-transformation analysis shows high rank stability (Spearman rho=0.986-0.999; Kendall tau=0.952 - 0.990). The results demonstrate that prescriptively admissible configurations can contain materially different implicit safety margins. The framework is intended as a transparent comparative screening method for regulatory analysis, not as an estimator of absolute fire-event probabilities.