Industries that fall under the legal requirements of the EU Seveso Directives must produce a safety report principally demonstrating that major-accident hazards have been identified and that the necessary measures have been taken to prevent their consequences. This legal specification involves the choice of a risk analysis approach, usually deterministic or probabilistic. These approaches have been formalized since the early of the 60's for the most of them, and they are based upon the consultation of expert judgments and reliability databases (for technical and human aspects). Due to the complexity of the industrial systems and their own dynamic in time and space, these risk assessment methods need to be supported by a systemic vision of their processes. Based on the systemic theory introduced by Forrester, this paper aims to propose a methodology to modeling and simulate the functioning of a chlorine storage and distribution unit, in order to promote a better risk prevention. This methodology include four complementary steps: i) the modeling of the system using the systemic theory principles and terms; ii) the implementation of this model into a Dynamic System platform for its simulation; iii) the simulation of the system using normal and abnormal scenarios in order to identify and to estimate the consequences of these deviations; iv) the assessment of the proposed barriers efficiency. After the application of this systemic methodology, these results are introduced into a HAZOP analyze to support the experts'risk assessment. As a conclusion, the authors discuss the benefits and limits of the implementation of a systemic approach in the safety reports, and propose some ways to generalize the proposed methodology.