Proceedings of the Eighth International Conference on Probabilistic Safety Assessment & Management (PSAM)
Risk-Informed Decision Making: Cost-Benefit
Download citation file:
- Ris (Zotero)
- Reference Manager
In the allocation of resources to system safety, there are two dominant approaches in the literature and practice: (i) multiobjective tradeoff analysis involving metrics of risk, performance, and cost; and (ii) cost-benefit analysis involving the monetization of benefits and risk reductions and costs. Multiobjective tradeoff analysis offers an advantage of bringing relevant evidence to the decision makers, who is thoroughly engaged in the evaluation of allocation alternatives. In addition, multiobjective tradeoff analysis is able to address a plurality of perspectives on the allocation problem. However the amount of information needing to be understood by the decision makers can be overwhelming, particularly when the decision space is combinatorial. Cost-benefit analysis offers an advantage of fiscal accountability in large safety programs, in that all benefits and risk reductions are monetized according to a set of standard assumptions. In addition, cost-benefit analysis effectively codifies the values of a safety program for uniform and consistent application across levels and divisions of an organization. However, it can be difficult to agree on the relevant perspectives of for example who pays and who benefits and furthermore to agree on the functional forms and parameters of the benefit and cost functions. We recognize the need for safety program managers to achieve the combined advantages and minimize the disadvantages of the several approaches. In this paper, we review recently developed methods to integrate multiobjective tradeoff analysis to cost-benefit analysis in the allocation of resources to system safety. We describe a framework of multiobjective combinatorial optimization that is supported by alternatives forms of the cost and benefit functions using interval analysis to address uncertainty of the function parameters. We describe that the integrated methods combine some advantages of bringing relevant evidence to the decision makers while supporting the fiscal accountability of the safety program. The methods illuminate a prevalent tradeoff between addressing the intensity of hazards and the exposures to the hazards. Impact analysis of the functional forms and interval analysis of uncertainty enable the methods to proceed without precise knowledge or agreement on key elements of cost and benefit functions. On a graphic representing the tradeoff analysis of risk, performance, and cost, we identify three zones of the plausibility of the benefit-to-cost ratio of particular allocations exceeding a needed threshold of the benefit-to-cost ratio: (1) certain to exceed the threshold, (2) plausible to exceed the threshold, and (3) impossible to exceed the threshold. There are applications of the methods to the allocation of safety-enhancing devices in transportation systems and to the allocation of biological, chemical, and radiological warning sensors around a region. We describe needs for future effort to improve the integration of multiobjective tradeoff analysis with cost-benefit analysis for decision making in safety programs. The paper will be of interest to program managers, reliability engineers, quality and safety professionals, and risk analysts who are charged with the allocation of limited investments and other resources to many and diverse kinds of technologies and projects in large safety programs.