A different way of thinking about seismic risk

Seismic risk has traditionally been approached using probabilistic analysis. This dilutes the potential impact of low probability, extreme events that may lead to severe consequences including excessive land damage, building damage, injuries and death. The communication of risk in probabilistic terms is also not clearly understood by most audiences. Further, it is evident that few building developers, owners and users have a good understanding the implications of this and the capacity design of buildings, which may not be repairable after a severe event. There is also an adverse impact on planning and land use, where decisions that may affect many people are based on a limited view of adverse outcomes such as liquefaction, lateral spread and slope stability in severe earthquakes. A different way of thinking about seismic risk is proposed. An approach of using scenarios derived from a combination of deterministic as well as probabilistic thinking would prompt consideration of impacts over a range of events. This would allow full consideration of which outcomes are clearly not acceptable and which are. This may facilitate planning for both private and public sector, with a common understanding that is relatively easily communicated to both experts and lay people. This risk evaluation framework would also facilitate consideration of mitigation, by bringing focus on unacceptable outcomes of severe events that are currently obscured by pure probabilistic analysis. This was missing in Christchurch, which experienced the sort of event we can readily anticipate and should actively plan for in other parts of New Zealand. This would help us avoid future red zones and excessive damage and demolition. It will inform development of building codes and standards and will help us evaluate risk and provide resilience and redundancy across the range of interconnected infrastructure networks. Informed debate is needed with key decision makers to discuss the underlying objectives of our regulation and how these may be better met by such an approach, without engineers allowing themselves to be trapped in past thinking and assumptions.