Multi-Objective Optimization of Internal Compartment Layout of Oil Tankers

This work deals with the design of the internal layout of a shuttle tanker formulated as a multi-objective optimization problem, balancing cargo capacity and minimizing still water bending moment with safety requirements, in particular survivability after damage. A parametric model is used to specify the internal layout of a tanker ship considering a fixed hull shape and regulatory framework. The design variables include positions of watertight members in the internal layout, such as watertight bulkhead position, double-bottom height, and wing tanks width. Merit functions are the minimization of oil outflow parameter, maximization of cargo capacity, and minimization of the longitudinal bending moment, which are, respectively, represented for reduction of environmental pollution due to damaged oil tankers, improvement of economic benefits, and safety during operation. The multi-objective genetic algorithm is used for approaching the Pareto frontiers, and the choices between the optimal designs are discussed while introducing a utility function. The internal layout of the ship's hull is established in the initial stage of ship design. The size and location of the internal spaces are defined taking into consideration the type of ship and the type of content and usage of the compartments. On the other hand, survivability regulations impose limitations in the subdivision arrangement of the ship. Also, the economic competitiveness has special importance in the design of a merchant ship. Cargo capacity and building cost are some of those economic objectives that are not necessarily in accordance with the improvement of safety objectives. Thus, the design of the internal layout of the vessels can be studied as a multi-objective optimization problem targeting safety improvement after damage besides decreasing the cost functions. A diversity of internal layout designs can be investigated within an optimization framework to choose the dominating design among the feasible solutions (Nowacki 2010). However, targeting a variety of objective functions complicates the process of internal compartment design, especially when the main hull dimensions have to be kept constant (Santos & Guedes Soares 2010).