Life Safety Risk Criteria in Seismic Decisions

This paper addresses acceptable life-safety risk within the context of riskmitigation decision-making. The gathering of risk information and risk-mitigation alternatives raises two questions: is a current level of risk acceptable, and if not, how does one determine if a particular mitigation measure is reasonable and prudent? A procedure is proposed for decision-makers to select among competing riskmanagement alternatives and to defend the selection. The procedure applies four tests to each alternative: necessity (is the existing risk unacceptably high); efficacy (is the alternative capable of mitigating risk to acceptable levels); efficiency (the relationship between cost and benefit of the risk-mitigation alternative); and how well the alternative provides for social equity. The necessity test relies on the decisionmaker’s subjective judgment of acceptable risk. Though no universally accepted measure of acceptable risk exists, guidance can be provided by legal and other precedent regarding the peril in question, and by considering levels of risk from related perils deemed acceptable by other authorities and decision-makers. The efficacy test examines whether a proposed mitigation alternative results in an acceptable level of risk, so its criteria are identical to the necessity test. The efficiency test can depend on subjective judgments of what is a reasonable cost for an incremental reduction in life-safety risk, or a reasonable cost per statistical life saved. Guidance on reasonable levels of efficiency can be found in legal and behavioral precedent, and by theoretical means using decision theory. The equity test relies on an examination of quantitative risk to people in the various social and economic groups affected by the mitigation decision. The necessity, efficacy, efficiency, and equity (NE) tests are illustrated with a case study that considers the decision to add seismic bracing to automatic sprinklers in highrise buildings. In the study, it is determined that the measure is justifiable in terms of the NE tests. It is noted that risk-management decisions are made within a broader context. Consequently, the NE tests may not be adequate to resolve fully the risk-management decision, but they can form a quantitative basis for informed discussion.