Improved estimation of ultra-deep water pipe collapse pressures by using extreme value theory

被引:3
作者
Heffernan, Janet E. [1 ]
Walker, Alastair [2 ]
Liu, Ping [3 ]
机构
[1] J Heffernan Consulting Ltd, Ulverston, England
[2] Adv Mech & Engn Pty Ltd, S Perth, WA, Australia
[3] INTECSEA BV, Delft, Netherlands
关键词
Bayesian methods; non-normal data; probability distributions; process improvement; product reliability; statistical modeling;
D O I
10.1080/08982112.2018.1482340
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
We examine safety margins between measured pipe collapse pressures and maximum external hydrostatic pressure exerted on pipeline on the seabed. We analyze collapse pressures measured on specimen rings cut from pipes, which are valid proxies for testing full length pipes. Our novel application of statistical models motivated by Extreme Value Theory provides a firmer basis for extrapolating collapse pressures than current practice which assumes a Gaussian distribution, and which gives over-conservative safety margins compared with our method. Our more accurate assessment of collapse pressures could give large cost savings in the manufacture of thick-walled pipe for ultra-deep water pipelines.
引用
收藏
页码:363 / 374
页数:12
相关论文
共 17 条
[1]   Interrelationship between statistical methods for estimating the size of the maximum inclusion in clean steels [J].
Anderson, CW ;
Shi, G ;
Atkinson, HV ;
Sellars, CM ;
Yates, JR .
ACTA MATERIALIA, 2003, 51 (08) :2331-2343
[2]  
[Anonymous], 2012, EXTREMES RELATED PRO
[3]  
[Anonymous], 2001, An Introduction to Statistical Modelling of Extreme Values
[4]  
DAVISON AC, 1990, J ROY STAT SOC B MET, V52, P393
[5]  
Dhanak M. R., 2016, Springer Handbook of Ocean Engineering
[6]  
DNVGL, 2013, SUBM PIP SYST
[7]  
Finkenstadt B., 2003, EXTREME VALUES FINAN
[8]   Pitting corrosion:: Comparison of treatments with extreme-value-distributed responses [J].
Fougères, AL ;
Holm, S ;
Rootzèn, H .
TECHNOMETRICS, 2006, 48 (02) :262-272
[9]  
Gilks W.R., 1995, CHAPMAN HALLCRC INTE
[10]   A comparison of the main methods for estimating probabilities of extreme still water levels [J].
Haigh, Ivan D. ;
Nicholls, Robert ;
Wells, Neil .
COASTAL ENGINEERING, 2010, 57 (09) :838-849