Benefit/Cost Ratio in Systems Engineering: Integrated Models, Tests, Design, and Production

Nitta, C ; Logan, R ; Chidester, S ; Foltz, M F

Lawrence Livermore National Laboratory

关键词: Probability;    Lawrence Livermore National Laboratory;    Testing;    Dollars;    37 Inorganic, Organic, Physical And Analytical Chemistry;   

DOI  :  10.2172/15011407
RP-ID  :  UCRL-TR-207610
RP-ID  :  W-7405-ENG-48
RP-ID  :  15011407

美国|英语

来源: UNT Digital Library

【 摘 要 】

We have previously described our methodology for quantification of risk and risk reduction, and the use of risk, quantified as a dollar value, in the Value Engineering and decision tradeoff process. In this work we extend our example theme of the safety of reactive materials during accidental impacts. We have begun to place the validation of our impact safety model into a systems engineering context. In that sense, we have made connections between the data and the trends in the data, our models of the impact safety process, and the implications regarding confidence levels and reliability based on given impact safety requirements. We have folded this information into a quantitative risk assessment, and shown the assessed risk reduction value of developing an even better model, with more model work or more experimental data or both. Since there is a cost incurred for either model improvement or testing, we have used a Benefit/Cost Ratio metric to quantify this, where Benefit is our quantification of assessed risk reduction, and cost is the cost of the new test data, code development, and model validation. This has left us with further questions posed for our evolving system engineering representation for impact safety and its implications. We had concluded that the Benefit/Cost Ratio for more model validation was high, but such improvement could take several paths. We show our progress along two such paths; simple and high fidelity modeling of the impact safety process, and the implications of our knowledge and assumptions of the probability distribution functions involved. At the other end of the systems engineering scale, we discuss the implications of our linkage from model validation to risk on our production plant operations. Naturally, the nature of most such methodologies is still evolving, and this work represents the views of the authors and not necessarily the views of Lawrence Livermore National Laboratory.