Impact of performance interdependencies on structural vulnerability: A systems perspective of storm surge risk to coastal residential communities

Interaction between residential structures during natural hazards can lead to interdependencies in their performance. During storm surge, for example, structures can affect the performance of inland buildings by generating damaging waterborne debris or by beneficially dampening surge loads. Quantifying the impact of this interaction on structural vulnerability is critical for risk assessment and informed decision-making. In this study we present and implement two general modeling approaches for investigating such interdependencies. The first method is to condition the vulnerability of a structure on the performance of neighboring buildings using a Markov model. The second uses a marginal model to account for correlation between damage observations when estimating a structure's vulnerability to the hazard. Both approaches are implemented using a case study of an impacted coastal community during Hurricane Sandy (2012). Findings indicate that a structure's performance during storm surge is strongly dependent on the damage state of the structure immediately seaward. Furthermore, considering the correlated damage states of buildings increases statistical uncertainty when relating structural performance to hazard intensity. Motivated by these findings, we propose a more coordinated approach to coastal risk mitigation which considers the effects of interdependencies on insurance pricing, structural design, mitigation strategies and community resilience. - Highlights: • Interaction between residential structures leads to performance interdependencies. • Interdependencies during storm surge are due to debris and structural shielding. • Markov model treats interdependencies as an additional demand parameter. • Marginal model incorporates damage correlation into regression estimation. • System behavior should be considered in community risk and resilience.