Study Of Risk Assessment And Control Models For Engine Room Fire Based On Life Cycle

With the development of marine transportation and exploitation of resources in response to the market requirment of ocean shipping and energy, the development of ships is towards large-scale, high-speed and specialization in order to expand the volume of sea cargo, and offshore platforms step gradually from shallow water to deep sea. Therefore, the field of risk in shipping and marine engineering has been given rise to more and more attention. Based on the research of the current life cycle risk analysis, ship and offshore platform life cycle risk analysis is discussed, and the contents include life cycle risk assessment and risk control models on engine room fire of ship and offshore platform.(1) Risk assessment of ship and offshore platform is a complicated system. The current studies mainly focus on a specific risk in ship and offshore platform life cycle. In this paper, life cycle risk assessment methods of ship and offshore platform are studied from the system engineering point of view.First of all, according to the characteristics, the life cycle of ships and offshore platforms can be divided into five phases, that is, planning, design, construction, operation and decommissioning. All the risks consist of five categories, that is, environmental risk, technical risk, economic risk, management risk and behavioral risk. Risk loss is graded according to the consequence. Then after risk identification for each phase, life cycle risk assessment index system of ships or offshore platforms can be established.In terms of the complexity and uncertainty involved in real world decision problems and how much information can be obtained, the expression of certainty or uncertainty are introduced into risk assessment methods, In the course of risk estimatement, judgments are given by experts. Because of the complexity of the practical problems and uncertainty of risks, experts are not always able to elicit exact judgments. If the problem is relatively simple, or has a certain amount of available information, experts can make judgment with crisp numbers, in which case the Grey-AHP method can be conducted for the life cycle risk assessment. If the problem is relatively complex, and the necessary information is not sufficient, it is unrealistic and infeasible to acquire exact judgments. Relatively it is easier for experts to use the upper and lower bounds for risk estimation, so interval number is taken to express experts' judgments. Interval number Grey-AHP method is proposed for life cycle risk assessment and satisfactory consistency and priority generation methods are studied. (2) In the field of ship and offshore platform risk control, most of current studies focus on the control methods at specific phase of life cycle. However, it is essential to study life cycle based risk control of ship and offshore platform.Life cycle risk control model of ship and offshore platform is established. Depending on the characteristics of various risks, possibility of risk control in each phase of life cycle is analyzed. In order to make decision for risk mitigation, risk control methods at different phases of life cycle are investigated. The analysis of the engine room fire demonstrates that the risk reduction plan can be taken at design and operation phases and the risk control methods of engine room fire is analyzed according to various incentives.In order to control the risk in engine room at design phase, large eddy simulation(LES) is adopted to analyze the effect of smoke diffusion and temperature distribution caused by ship structure, and diffusion law of fire smoke under deck is obtained. On this basis, smoke pool effect is proposed and its characteristics are analyzed. As an example, fire scenario in engine room of a 26000DWT tanker is simulated in conditions of three different ventilations. According to the simulation results, suggestion is given with regard to ship design.In order to control the risk in engine room at operation phase, Bayesian network model of engine room fire is built. For the purpose of finding out the situation after the engine room fire occured, manual fire detection is often implemented. It is not only dangerous, but also time-consuming, so the best time of extinguishing is often lost. To solve the problem, Bayesian network model of engine room fire can be used for reasoning and providing reference for firefighting decision.