Oil Tanker Design Methodology Considering Probabilis- Tic Accident Damage

This paper proposes a total-system design methodology that includes four important components necessary for a systematic approach to oil tanker concept design. These are: • An efficient and effective search of design space for optimal or non-dominated designs • Well-defined and quantitative measures of objective attributes, cost and risk • An effective format to describe the design space and to present non-dominated concepts for rational selection by the customer • A probabilistic method for predicting damage in tanker accidents and improving tanker crashworthiness A Multi-Objective Genetic Optimization (MOGO) is used to search design parameter space and identify non-dominated design concepts based on total ownership cost and oil outflow risk. A simplified ship synthesis model balances the designs and assesses their feasibility. A cost model estimates total ownership cost which is defined to include construction cost, and discounted operational and maintenance costs. A risk model calculates the probability of grounding and collision and the resulting oil outflow in bottom and side damage. Damage is predicted using simplified damage models in a Monte Carlo simulation. The optimization considers hull form characteristics, producibility, cargo block subdivision and double hull dimensions, structural design characteristics, propulsion and power redundancy, manning and automation, and trade route and port characteristics. The emphasis of this paper is on process. Some preliminary results are presented.