A decision model for the risk management of hazardous processes

    Research output: Book/ReportReport


    A decision model for risk management of hazardous processes as an optimisation problem of a point process is formulated in the study. In the approach, the decisions made by the management are divided into three categories: (1) planned process lifetime, (2) selection of the design and, (3) operational decisions. These three controlling methods play quite different roles in the practical risk management, which is also reflected in our approach. The optimisation of the process lifetime is related to the licensing problem of the process. It provides a boundary condition for a feasible utility function that is used as the actual objective function, i.e., maximizing the process lifetime utility. By design modifications, the management can affect the inherent accident hazard rate of the process. This is usually a discrete optimisation task. The study particularly concentrates upon the optimisation of the operational strategies given a certain design and licensing time. This is done by a dynamic risk model (marked point process model) representing the stochastic process of events observable or unobservable to the decision maker. An optimal long term control variable guiding the selection of operational alternatives in short term problems is studied. The optimisation problem is solved by the stochastic quasi-gradient procedure. The approach is illustrated by a case study.
    Original languageEnglish
    Place of PublicationEspoo
    PublisherHelsinki University of Technology
    Number of pages23
    ISBN (Print)978-951-22-3511-7
    Publication statusPublished - 1997
    MoE publication typeD4 Published development or research report or study

    Publication series

    SeriesHelsinki University of Technology: Systems Analysis Laboratory. A: Research Reports


    Dive into the research topics of 'A decision model for the risk management of hazardous processes'. Together they form a unique fingerprint.

    Cite this