Page 296 - Practical Design Ships and Floating Structures
P. 296

27 1

        assessed on a comparative basis, applying the consequence analysis tools and the risk-based techniques,
        including methods to analyse the profitsllosses involved. Alternative design solutions will therefore be
        suggested, their selection being  based  on  a  sound economic  application of the QRA-based  design
        procedure.


        3  ELEMENTS OF THE PROCEDURE

        All  design  decision models  developed  so  far,  are  based  on  a  single economic criterion of  merit,
        incorporating  several constraints related  to performance  indicators,  lately with the addition of some
        safety indicators of any type (deterministic, probabilistic or performance-based).  In the context of the
        above described design goal, the current approaches  are not sufficient, due to the fact that safety is not
        an integral part of the design process, but is taken into consideration  as a design periphery issue at best,
        if not as a design afterthought. For this reason, the association of increased safety leading to increased
        incurred costs is considered to be the norm.

        A clear and complete statement of the goal for the problem at hand is that a design procedure is sought
        to “derive effective  arrangements and layouts that  maximise safety, whilst minimising the  incurred
        costs”.  This is a multiple criteria problem, the design solution of which depends on multiple design
        attributes, which in turn derive from the attained performance and characteristics of the alternatives
        under consideration.

        To  achieve such  a goal a  structured formulation  of the criteria, parameters,  constraints, objective
        function and mathematical models, needs to be developed.  This will be based on the observation that
        an alternative to a single criterion of merit is the consideration of pair wise comparisons that employ
        valid  criteria, which can also be  extended to the consideration of a hierarchically decompositioned
        objective function that  reflects and  combines  economic,  performance  and  safety aspects.  Such  a
        formulation will lead to the following innovative aspects:

           The criteria to be considered need only be design-related, and not of conformance nature.  More
           importantly,  these criteria can be  incorporated  in the  formulation of the objective function,  as
           described below.
           The formulation allows for the development of a practically unconstrainted problem, in the sense
           that the various indicators (economic, performance, safety) are included in the objective function.
           The mathematical models ought to bc of pcrformance nature, at least for the assessment of damage
           survivability. This will allow the incorporation of first-principles approaches in the design process,
           which  when  combined  with  an  overall  QRA  framework,  accounts  for  an  objective  safety
           quantification process.

        In the following, the formulation of the decision  making process for the determination of the most
        effective internal  configuration, given the vessel’s  hull,  will  be  described.  The  focus will be  on
        monohull passenger Ro-Ro vessels.

        3.1  Development of the Objective Function
        The objective function will be developed based on a method proposed by  T.L.  Saaty, known as the
        Analytic  Hierarchy  Process  (AHP),  Saaty  (1980).   The  method  consists  of  three  principles:
        decomposition, comparative judgement, and priority synthesis.
        Decomposition is the description of the problem in a hierarchical  form.  The elements of each level are
        independent of  succeeding levels.  The hierarchical  structure starts at the top with a statement of  a
        decision goal.  The next lower levels contain the criteria by which the alternatives are measured by the
   291   292   293   294   295   296   297   298   299   300   301