80% of road travel out of Wellington wouldn't be possible after 7.5 magnitude quake - study

A study simulating what would happen to Wellington's road network in a major earthquake has found some parts would be cut off for months.

The simulation assessed the disruption that would be caused by a 7.5 magnitude quake on the Wellington Fault.

The researchers referenced a 2011 study that estimated a 10 percent chance such an earthquake would strike in the next 100 years.

They simulated traffic using parameters like driver reaction time and jam density then applied it to a map of the North Island based on a previous study assessing the infrastructure damage caused by such an earthquake.

On the map, several key roads had been closed including multiple stretches of State Highway 1, Hutt Road, Grant Road, Curtis Street and Chaytor Street.

In the immediate three-day aftermath, the researchers found 68 percent of trips into the Wellington region and 82 percent of trips out of Wellington would be infeasible.

After three months, that reduced to 25 percent of trips into Wellington and 49 percent of trips out of Wellington.

They found Lower Hutt, followed by Wellington City, would be most impacted.

Researchers also found successful trips were much slower, with simulated traffic spreading onto urban detour routes which increased travel times.

But the researchers acknowledged the simulation was limited by assuming demand for travel would remain steady, while in reality many motorists would baulk at the travel time and decide not to drive.

"While this assumption enabled stress-testing of network performance under disruption, it may overestimate congestion and trip cancellations by not accounting for behavioural adaptation during recovery," the paper said.

The researchers said the results highlighted the importance of spending on critical roads to ensure they stayed resilient.

"Overall, the findings highlight that proactive planning, demand management, and targeted investment in critical corridors are essential to maintain network functionality and enhance resilience under extreme but plausible post-disaster scenarios," they wrote.