Abstract
| Original language | English |
|---|---|
| Article number | 012102 |
| Journal | Journal of Physics: Conference Series |
| Volume | 364 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 2012 |
| MoE publication type | A1 Journal article-refereed |
| Event | 25th International Congress on Condition Monitoring and Diagnostic Engineering, COMADEM 2012 - Huddersfield, United Kingdom Duration: 18 Jun 2012 → 20 Jun 2012 Conference number: 25 |
Fingerprint
Keywords
- Decision support
- failure list
- fault tree analysis
- maintenance
- online
- probability
Cite this
}
Fault tree analysis for maintenance needs. / Halme, Jari; Aikala, Antti.
In: Journal of Physics: Conference Series, Vol. 364, No. 1, 012102, 2012.Research output: Contribution to journal › Article › Scientific › peer-review
TY - JOUR
T1 - Fault tree analysis for maintenance needs
AU - Halme, Jari
AU - Aikala, Antti
PY - 2012
Y1 - 2012
N2 - One of the key issues in maintenance is to allocate focus and resources to those components and subsystems which are the most unreliable and prone to failures. In industrial systems, fault tree analysis technique can be used to study the reliability of the complex systems and their substructures. In this paper a fault tree application for analyzing online the current reliability and failure probability for maintenance purposes is presented. The analysis is utilizing data connected to the fault tree root causes and events. An indication of an anomaly case, service action, cumulative loading, etc., or just time passed or service hour counter level can trigger a new calculation of current probabilities of the fault tree events and subsystem interactions. In proposed approach real time, dynamic information from several available data sources and different measurement are interconnected to each fault tree event and root cause. There is also formulated an active, constantly updated link between the fault tree events and maintenance databases for the maintenance decision support, and to keep the analysis up to date. Typically top event probability is evaluated based on updated root cause probabilities and lower level events. At the industrial plant level an identification of a failure in a component event defined within a constructed and operatively existing fault tree explicitly means that the event's failure probability is one. By utilizing this indication, the most probable failure branches through the fault tree sub events to root causes can be identified and printed as a valid check list for maintenance purposes to focus service actions first to those fault tree branches most probable causing the failure. Respectively, during the checks, service actions, etc., components, especially those within the critical branches, detected as healthy can be a updated as having zero failure probability. This information can be used to further update the fault tree and produce online a new service order list. The added value of the proposed method with respect to developed software platform functions lies in its applicability to rationalize maintenance actions and in a case of a failure allocate resources where they are assumable mostly needed.
AB - One of the key issues in maintenance is to allocate focus and resources to those components and subsystems which are the most unreliable and prone to failures. In industrial systems, fault tree analysis technique can be used to study the reliability of the complex systems and their substructures. In this paper a fault tree application for analyzing online the current reliability and failure probability for maintenance purposes is presented. The analysis is utilizing data connected to the fault tree root causes and events. An indication of an anomaly case, service action, cumulative loading, etc., or just time passed or service hour counter level can trigger a new calculation of current probabilities of the fault tree events and subsystem interactions. In proposed approach real time, dynamic information from several available data sources and different measurement are interconnected to each fault tree event and root cause. There is also formulated an active, constantly updated link between the fault tree events and maintenance databases for the maintenance decision support, and to keep the analysis up to date. Typically top event probability is evaluated based on updated root cause probabilities and lower level events. At the industrial plant level an identification of a failure in a component event defined within a constructed and operatively existing fault tree explicitly means that the event's failure probability is one. By utilizing this indication, the most probable failure branches through the fault tree sub events to root causes can be identified and printed as a valid check list for maintenance purposes to focus service actions first to those fault tree branches most probable causing the failure. Respectively, during the checks, service actions, etc., components, especially those within the critical branches, detected as healthy can be a updated as having zero failure probability. This information can be used to further update the fault tree and produce online a new service order list. The added value of the proposed method with respect to developed software platform functions lies in its applicability to rationalize maintenance actions and in a case of a failure allocate resources where they are assumable mostly needed.
KW - Decision support
KW - failure list
KW - fault tree analysis
KW - maintenance
KW - online
KW - probability
U2 - 10.1088/1742-6596/364/1/012102
DO - 10.1088/1742-6596/364/1/012102
M3 - Article
VL - 364
JO - Journal of Physics: Conference Series
JF - Journal of Physics: Conference Series
SN - 1742-6588
IS - 1
M1 - 012102
ER -