You need JavaScript to view this

A Bayesian inspection procedure applied to offshore steel platforms

Technical Report:

Abstract

Offshore structures are susceptible to fatigue damage due to the time varying hydrodynamic loads and the high stress levels. Therefore, in order to obtain a given high reliability of the structure, inspection schemes have to be formulated in order to keep track of crack initiation and crack growth. Usually the inspections and the subsequent repairs have to be done on-site, requiring a skilled work-force and expensive man-hours. It is therefore important to optimize the intervals between the inspections, the number of joints inspected and the amount of repair work. In the present paper a Bayesian approach is taken, following a procedure originally suggested by Itagaki and Shinozuka. Here the statistical distributions of some of the parameters in the crack initiation, crack detection and crack propogation are updated at each inspection and used to derive better estimates for the number of severe cracks as function of time until the next inspection. Thereby it is possible to decide on an appropriate interval to the next inspection taking into account both the cost of the inspection and the cost of repairing the severe cracks. As an example, a jack-up operating in the North Sea is chosen. Inspection data for 3 consecutive inspections of  More>>
Publication Date:
Apr 01, 1992
Product Type:
Technical Report
Report Number:
DTH-DCAMM-441
Reference Number:
SCA: 423000; PA: DK-92:001815; SN: 93000917920
Resource Relation:
Other Information: PBD: Apr 1992
Subject:
42 ENGINEERING; OFFSHORE PLATFORMS; SYSTEM FAILURE ANALYSIS; FATIGUE; MAINTENANCE; CRACKS; INSPECTION; CRACK PROPAGATION; PROBABILISTIC ESTIMATION; MATHEMATICAL MODELS; 423000; MARINE ENGINEERING
OSTI ID:
10110757
Research Organizations:
Technical Univ. of Denmark, Lyngby (Denmark). Inst. of Mathematical and Operations Research
Country of Origin:
Denmark
Language:
English
Other Identifying Numbers:
Journal ID: ISSN 0106-6366; Other: ON: DE93752739; TRN: DK9201815
Availability:
OSTI; NTIS
Submitting Site:
DK
Size:
14 p.
Announcement Date:
Jun 30, 2005

Technical Report:

Citation Formats

Juncher Jensen, J, and Terndrup Pedersen, P. A Bayesian inspection procedure applied to offshore steel platforms. Denmark: N. p., 1992. Web.
Juncher Jensen, J, & Terndrup Pedersen, P. A Bayesian inspection procedure applied to offshore steel platforms. Denmark.
Juncher Jensen, J, and Terndrup Pedersen, P. 1992. "A Bayesian inspection procedure applied to offshore steel platforms." Denmark.
@misc{etde_10110757,
title = {A Bayesian inspection procedure applied to offshore steel platforms}
author = {Juncher Jensen, J, and Terndrup Pedersen, P}
abstractNote = {Offshore structures are susceptible to fatigue damage due to the time varying hydrodynamic loads and the high stress levels. Therefore, in order to obtain a given high reliability of the structure, inspection schemes have to be formulated in order to keep track of crack initiation and crack growth. Usually the inspections and the subsequent repairs have to be done on-site, requiring a skilled work-force and expensive man-hours. It is therefore important to optimize the intervals between the inspections, the number of joints inspected and the amount of repair work. In the present paper a Bayesian approach is taken, following a procedure originally suggested by Itagaki and Shinozuka. Here the statistical distributions of some of the parameters in the crack initiation, crack detection and crack propogation are updated at each inspection and used to derive better estimates for the number of severe cracks as function of time until the next inspection. Thereby it is possible to decide on an appropriate interval to the next inspection taking into account both the cost of the inspection and the cost of repairing the severe cracks. As an example, a jack-up operating in the North Sea is chosen. Inspection data for 3 consecutive inspections of several joints close to the lower guides are available. Based on an analysis of these results the implementation and usefulness of the present Bayesian approach is discussed. (Author).}
place = {Denmark}
year = {1992}
month = {Apr}
}