Sensitivity of Optimal Decision Outcome to Consequence Assumptions in Value of Information Analysis

Abstract

Value of information is directly related to the expected reduction in cost of con-sequences due to the availability of the inspection information. However, the estimation of con-sequence cost can be highly variable due to consequences that are difficult to quantify such as safety costs, reputational costs, regulatory costs, and impact on societal license for future business development. While the intangible costs are difficult to quantify with precision, the range of possible costs can be estimated from previous events and the other business cost estimates. Inclusion of this possible range for cost of consequences, instead of single value of best estimate, provides additional insight for decision-making using the value of information analysis. This study presents the approach for a buried natural gas transmission pipeline. The cost paradigm considers the failure consequence costs at a component level and system level for the transmission network, and impact of the uncertainty in the failure costs due to the scale of the system considered is presented. A numerical example is presented to demonstrate the decision-outcomes with consideration of the cost uncertainty.