Abstract

Dynamic thermodynamic calculation data of sea ice thickness since 1947 to 1996 in Chengdao oilfield is applied here to calculate corresponding design ice thickness. The fitting curves for these observations are selected from Gumbel, Weibull, lognormal, Pearson type-3 and maximum entropy distribution. Corresponding return values given by these curves are regarded as the best design sea ice thickness parameters. Based on distribution fitting tests and comprehensive consideration, lognormal distribution is chosen as the best fitting curve of annual extreme data for sea ice thickness conclusively. Then different return values can be deduced under different return periods, and maximum likelihood method is applied to determine interval estimations of these return values. The calculated results can provide a reference for disaster prevention and offshore structures design.

Keywords:
maximum entropy distribution, ice thickness, artificial intelligence, confidence interval, liaodong bay, weibull distribution, calculation, gumbel distribution, sea ice thickness, pearson type-3 distribution
Subjects:
Information Management and Systems
This content is only available via PDF.
2014. International Society of Offshore and Polar Engineers
You can access this article if you purchase or spend a download.