Towards a novel design perspective for system vulnerability using a Markov chain

Abstract

In order to investigate to which extent naval ships can execute their operational scenario after damage, an early stage assessment of the vulnerability of distributed systems needs to be carried out. Such assessments are currently mostly done by evaluating the performance of predefined concepts. However, such an approach does not necessarily lead to the most desirable solution, since solutions outside the scope of the designer’s preconceived ideas or experience are inherently hard to investigate. This paper therefore proposes several steps towards an approach that enables a vulnerability assessment that is independent of predefined concepts. This is done by incorporating several additions to an existing system vulnerability approach developed by the authors, using a Markov chain. With this approach there is no longer a need for modelling individual hits or damage scenarios. Whereas the approach has previously been shown in concept, this paper introduces three improvements that contribute to the applicability of the approach: 1)it is scaled up in order to model a larger number of compartments and distributed systems, 2) the hit probabilities for different compartments can be adjusted, and 3) it is shown how the availability of main ship functions can be derived from the availability of individual connections. A test case that compares two powering concepts (conventional and full electric powering) of a notional Oceangoing Patrol Vessel (OPV) is provided to illustrate the principles behind the improvements. From the results the two main contributions of this paper can be obtained: 1)the possibility to assess the system vulnerability for different levels of required residual capacity at different impact levels, and 2) and the quantitative nature of the results, aiding ship designers and naval staff with understanding the consequences of various concepts on the system vulnerability.