Abstract
Infrastructure is frequently subject to various loading and environmental stressors that cause degradation of its performance with time. Management of such degrading infrastructure is conditioned upon its estimated performance. The conventional methodologies of infrastructure
performance assessment, such as the rating factor assessment, employ semi-probabilistic (or deterministic) techniques. It is known that the problem of infrastructure performance assessment is subject to various errors and uncertainties. Therefore, probabilistic approaches such as structural reliability assessment are recommended for performance assessment due to their ability to quantify and incorporate uncertainties of infrastructure performance. However, such probabilistic assessments require additional time and monetary costs while their potential monetary benefits are not apparent to the asset manager. In this article, the authors present a Bayesian decision framework and methodology to quantify the potential monetary benefits of probabilistic assessments. The framework is based on the value of the information (VoI) framework. The prior analysis focuses on infrastructure management using a semi-probabilistic code-based rating factor assessment, and the preposterior analysis focuses on
a reliability-based probabilistic assessment using a proposed conditional distribution of reliabilities. It is found that a probabilistic assessment can have significant benefits relative to a load rating factor assessment. A comparison of the conditional distribution of reliabilities with in-situ reliability estimates reveals the adequacy of the distribution. Infrastructure asset managers can utilize this framework to decide on probabilistic assessments over rating factor assessments.
performance assessment, such as the rating factor assessment, employ semi-probabilistic (or deterministic) techniques. It is known that the problem of infrastructure performance assessment is subject to various errors and uncertainties. Therefore, probabilistic approaches such as structural reliability assessment are recommended for performance assessment due to their ability to quantify and incorporate uncertainties of infrastructure performance. However, such probabilistic assessments require additional time and monetary costs while their potential monetary benefits are not apparent to the asset manager. In this article, the authors present a Bayesian decision framework and methodology to quantify the potential monetary benefits of probabilistic assessments. The framework is based on the value of the information (VoI) framework. The prior analysis focuses on infrastructure management using a semi-probabilistic code-based rating factor assessment, and the preposterior analysis focuses on
a reliability-based probabilistic assessment using a proposed conditional distribution of reliabilities. It is found that a probabilistic assessment can have significant benefits relative to a load rating factor assessment. A comparison of the conditional distribution of reliabilities with in-situ reliability estimates reveals the adequacy of the distribution. Infrastructure asset managers can utilize this framework to decide on probabilistic assessments over rating factor assessments.
Original language | English |
---|---|
Title of host publication | 14th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP14 |
Place of Publication | Dublin Ireland |
Publisher | Trinity College |
Number of pages | 8 |
Publication status | Published - 2023 |
Event | International Conference on Application of Statistics and Probability in Civil Engineering 2023 - Dublin, Ireland Duration: 9 Jul 2023 → 13 Jul 2023 Conference number: 14th https://icasp14.com (Website) http://www.tara.tcd.ie/handle/2262/102897 (Proceedings) |
Conference
Conference | International Conference on Application of Statistics and Probability in Civil Engineering 2023 |
---|---|
Abbreviated title | ICASP 2024 |
Country/Territory | Ireland |
City | Dublin |
Period | 9/07/23 → 13/07/23 |
Internet address |
|