Hierarchical Bayesian Reliability Analysis of Binomial Distribution based on Zero-Failure Data

The application of reliability-based structural integrity enables the process of quantitative risk assessment as part of pipelines’ integrity management program (IMP). This paper explores two topics that present challenges in terms of the practical adoption of a reliability-based IMP. The first challenge is the balance between perceived and true risk when implementing a quantitative reliability-based integrity model. This is a cornerstone for building stakeholder confidence in the calculated probability of failure (PoF) which is applied to safety and economically driven integrity decisions. The second challenge is the assurance that all relevant sources of uncertainty have been incorporated, which is essential for ensuring an accurate representation of the risk of failure of the pipeline. The level of conservatism (i.e. sufficient margin of error to maintain safety) incorporated when addressing these challenges may create a situation where calculated PoFs become inflated; becoming disproportionate to the failure history and contradictory to the current safe operation of pipelines being modeled. Two different PoF calibration approaches are proposed as practical options to address these challenges. The first method calibrates model error using an operator’s in-service failure history (i.e. failures that occurred under normal operation). The second method uses a set of failure data (including hydrostatic test failures and in-service failures) as selected by the operator considering key factors to ensure adequate representation of their specific pipeline system. These options will be demonstrated by assessing the integrity reliability of a hypothetical pipeline system. This work is expected to help evaluate the feasibility of challenging current practices regarding practical inclusion of epistemic uncertainty in integrity reliability analysis of pipelines.

Download Full-text

Supply and failure data reliability analysis system

Microelectronics Reliability ◽

10.1016/0026-2714(87)90216-2 ◽

1987 ◽

Vol 27 (2) ◽

pp. 382

Keyword(s):

Reliability Analysis ◽

Data Reliability ◽

Failure Data ◽

Analysis System

Download Full-text

Estimation of Failure Rate and its Applications in the Case of Zero-Failure Data

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.945-949.1046 ◽

2014 ◽

Vol 945-949 ◽

pp. 1046-1049

Author(s):

Ming Han

Keyword(s):

Bayesian Estimation ◽

Failure Rate ◽

Estimation Method ◽

New Method ◽

Hierarchical Bayesian ◽

Failure Data ◽

Hierarchical Bayesian Estimation ◽

Definition Of

This paper introduces a new method, named E-Bayesian estimation method, to estimate failure rate in zero-failure data. The definition of E-Bayesian estimation of the failure rate is given, based on the definition, the formulas of E-Bayesian estimation and hierarchical Bayesian estimation of the failure rate were provided, and properties of the E-Bayesian estimation, i. e. relations between E-Bayesian estimation and hierarchical Bayesian estimation, was discussed. Calculations were performed on practical problems, showing that the proposed new method is feasible and easy to operate.

Download Full-text

Nonparametric Reliability Analysis of Spacecraft Failure Data

Spacecraft Reliability and Multi-State Failures ◽

10.1002/9781119994077.ch2 ◽

2011 ◽

pp. 9-22

Keyword(s):

Reliability Analysis ◽

Failure Data

Download Full-text

Case Study on Reliability of Automotive Painting Shop Robots

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.1125.516 ◽

2015 ◽

Vol 1125 ◽

pp. 516-520

Author(s):

Mohd Amin Abd Majid ◽

Ahmad Fauzi Fudzin

Keyword(s):

Reliability Analysis ◽

High Reliability ◽

Mean Time Between Failures ◽

Assembly Plant ◽

Failure Data ◽

Time Between Failures ◽

Plant Maintenance ◽

Automotive Assembly ◽

Mean Time

Robots in automotive assembly plant are crucial for automation of the plant. As the robots are expensive, they are the main contributor to plant investment. Thus, it is important for the robots to have high reliability. This paper presents the reliability analysis of robots at an automotive assembly plant. The analysis was based on operating hours and frequency of the failures of the robots. The data were acquired from plant maintenance data. Using seven years of failure data, the robots’ mean time between failures (MTBF) and reliability were evaluated. Constant failure rate was assumed in the analysis. From the analysis it is noted that the reliability of the robots at the plant varies from 15% to 62% and 2% to 39% for 10,000 hours and 20,000 hours respectively. These findings could assist the maintenance manager to schedule the maintenance and replacement of the robot at the plant.

Download Full-text