Risk Assessment Of Failure Modes Related To Marine Boiler Using Fuzzy Expert System

Steam generation,in terms of marine boiler operation,involves a large number of potential failures and hazards,which impose high risks,causing deadly and costly accidents resulting in human casualties,machinery damage,ship operational disruption,and other financial losses.Boiler failure onboard merchant ships,leads to fatal and catastrophic consequences,is an onerous issue and involves critical risks especially for Marine machinery manufacturers,marine engineers and other related personal involved in its operation and maintenance.The identification of critical risks is essential to evaluate the safety,maintainability,and environment,to increase the reliability of the marine machinery for the ship operating companies,both the private and the government,academic institutions involved in the training and education of marine engineers and ratings.In this thesis,the 30 real-time accident-based potential failure modes related to the marine boiler,affecting the safety of merchant ships of different types,were summarized at first,by reviewing the past accident data and the reports of maritime transport and safety boards.Later on,for the accident-based risk assessment study with the help of experts’ consultations,elicitations using a technical questionnaire,opinions,and judgements from marine engineering experts were carried out to obtain the Occurrence(O),Severity(S),and Non-Detection(D)data for the 30 potential failure modes.Due to the ambiguity and uncertainty involved in the data of the O,S,and D,from the elicitations,where each of the O,S,and D,is presented on a 10-point scale;a fuzzy set approach is developed based on a rule-based fuzzy logic system for the critical risk evaluation,quantification,and prioritization.Assessment of the risk values for the 30 potential failure modes is performed by applying the fuzzy set approach to the data from the elicitations.It is found that among the 30 potential failure modes,“Fuel/Gas pipe Leakage" has the highest risk value,followed by “Active local pitting corrosion from water/Steam side",and “Soot fire",respectively.During the assessment,three different types of membership functions are employed to observe the effects of membership function on the corresponding risk values.Gaussian membership function is found more suitable than the triangular and trapezoidal membership functions.Finally,sensitivity analyses are performed for the potential failure mode with the highest risk value obtained from the risk assessments to show the dependence of the uncertainty of output fuzzy risk value on the uncertainty involved in the input values.It is expected that the results derived from this study can be used as a reference to help in the eradication of the potential failures and their adverse outcomes,ultimately contributing to the reliability of marine boiler system towards improving the ship safety.Thus,the knowledge obtained from this study can be applied for the benefit and further accident-based education of ship-crew,boiler operators onshore as well as offshore facilities,ship operating companies,machinery manufacturers,ship classification societies,owners,State Maritime authorities,policy-makers,shipyards and customers related to boiler service companies.The risk analysis methods developed in this study are believed to be applicable and useful in quantitative analysis for the risk assessment of marine machinery systems.