After thoroughly examined the critical machinery failures that occurred on the ships registered with NIPPON KAIJI KYOKAI (ClassNK) over the past 20 years, the proposed procedure for reliability evaluation of propulsion system was found practicable and effective. The procedure employed a general Failure Mode and Effect Analysis (FMEA) method in combination with so-called cause effect analysis for determining important failure modes, and utilised Reliability Block Diagram (RBD)-based computer software to calculate the reliability and availability of a system quantitatively.


Current ship classification rules largely consist of structural requirements for determining the dimensions of individual components based on single failure mode usually related to strength. However, most of failures experienced onboard occurred in unexpected modes. In addition, present rules are said to be lack of holistic approaches toward overall system reliability as well as safety, while much concentrating upon strength analysis for each of parts subjectively thought important. Based on such perception, the authors have been exploring the possibility of introducing reliability-based classification criteria in an effort to enable rules to cover multiple failure modes and evaluate a system as a whole both qualitatively and quantitatively. Although reliability methodology itself is not a new concept, its introduction to classification rules could be a milestone in terms of offering industry new ways of thinking and new tools for preventing and mitigating risk. After thoroughly examined the all critical machinery failures onboard the ships registered with ClassNK over the past 20 years, the authors found that practical reliability criteria for ship propulsion system could be set, taking into account the currently achieved reliability level. Also learned was that using the general FMEA method in combination with so-called cause effect analysis through actual failures could provide an efficient way in finding out major failure modes and weak points of a system.

This content is only available via PDF.
You can access this article if you purchase or spend a download.