UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Learning from failure : development and discussion of a database of structural failures Epp, Peter Yvon


This thesis presents and examines a database of structural failures, including bridges, buildings, and telecommunication towers. In addition to the database, a novelty of this work is the comparison of observed failure probabilities with target reliability index values in design codes. The study aims to improve the safety of future construction by exposing and understanding past structural failures. A labelling system is established to categorize and provide a basis for the future documentation of failures. The 275 failures in the database are collected from a search of newspaper articles, case studies, and the Imhof (2021) database of bridge failures. The scope of structural failures in this research considers partial and total collapse. The focus of the database is unexpected failures; therefore, failures due to vandalism, terrorism, or earthquakes are not considered. Earthquakes are excluded from the natural hazards considered because substantial damage is often expected during design-level ground shaking. Other natural hazards, such as ice accumulation, wind, and flooding, are considered. The majority of the failures that are considered in the database occurred in the United States; those account for 41% of the failures. Several insights are obtained from the total database. For instance, the most significant cause of failure for bridges is natural hazards, accounting for 29% of the failures. That is followed by 20% design errors and 15% failures caused by collisions. The most significant causes of failure for buildings are construction errors (24%), design errors (20%), and natural hazards (15%). The most significant failure causes for telecommunication towers are natural hazards (61%) and negligent maintenance errors (26%). The failures that occur during maintenance operations are primarily associated with the removal of structural members without adequate temporary support. The bridge superstructure type with the most failures is beam bridges. In the United States, truss bridges are apparently associated with the largest probability of failure. That failure probability corresponds to a reliability index of 3.1, which is inferior to the lifetime target reliability index of 3.5 from CSA S6.1.19.

Item Media

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International