Earthquake damage to a bridge in Kobe, Japan. Credit Damon Lee

+ Infrastructure design in low-seismicity areas should consider the ground movement that could occur in even the most unlikely earthquake.

Just because earthquakes are unlikely in areas of lower seismicity, it doesn’t mean cities like Sydney should be complacent and not design infrastructure for resilience in case of a rare event. I think this is a lesson that we should learn from the 5.6-magnitude earthquake that struck Newcastle, Australia, 25 years ago this month, killing 13 people and costing $A4 billion dollars.

A community whose infrastructure is still in working order after a major earthquake can recover faster. As infrastructure is typically founded in the ground, understanding ground effects (such as liquefaction) are critical to designing for resilience. In areas of high seismicity, the design of infrastructure explicitly considers ground impact. But what happens elsewhere?

Modern seismic design in high-seismicity areas is based on a maximum considered earthquake event generally determined by probabilistic methods. But we know that nature is unpredictable and the unlikely can happen – as it did recently in Japan and Christchurch, where the ground shaking exceeded the maximum considered design events. As a result, critical infrastructure was damaged and hindered the recovery.

In areas where you don’t expect earthquakes, such as Australia, considering these rare events becomes even more difficult.

What should be the maximum considered design in such areas? The inherent limited earthquake record in low-seismicity areas makes it difficult to establish design ground motions based on probabilistic methods without significant uncertainties.

In low-seismicity areas, infrastructure such as water mains, sewers, and bridges, aren’t considered lifelines and so aren’t governed by the maximum considered earthquake.

Yet would society accept damage to a bridge or water main that may take months or even years to repair? What would the impact on its community be? And what would the economic impact be? The long-term cost and society implications for the wider community aren’t used in the risk assessment methodology used to make these decisions.

The cost of mitigating the potential impact of these rare events is thought to be too high. But perhaps this is biased from a lack of earthquake engineering design experience in normal practice.

A resilient design that could withstand even rare events doesn’t necessarily have to cost much more. First, you have to understand the risk.