A catastrophic Alpine Fault earthquake could leave the South Island's West Coast cut off from the main grid for many weeks - and now researchers have begun exploring how the region could survive by itself.
The major fault, straddling the spine of the South Island, poses one of New Zealand's biggest natural disaster threats.
It's estimated to generate quakes of magnitude 7.5 or larger every three centuries and recent research gave a 29 per cent chance of a big "surface rupturing" event striking within the next 50 years.
Alongside work looking at what might happen in a magnitude 8-plus earthquake, known as Project AF8, an Auckland University researcher Associate Professor Nirmal Nair is investigating how tens of thousands of homes could cope with being cut off from the main grid for as long as two months.
In a new $234,000 two-year study, Nair is working alongside Westpower, which has a distribution network supplying more than 13,000 customers, to map out a potential "micro-grid" that could operate independently if forced to.
"There's a sense that electricity and energy infrastructure is highly reliable - and that's true - but no one has really studied the resilience aspect of it, and there's no metrics around this, especially in very large disasters."
In the case of an Alpine Fault quake, the loss of connectivity from the Transpower lines coming from the east could largely cut power to the West Coast, with the exception of alternate lines coming in from the north.
"So the scenario we have co-created, just for a case study, is that it would be six to eight weeks before electricity could be connected back to the main grid.
"And if this highly improbable event does happen, how do you operate in that?"
Nair estimated electricity demand, from Nelson south to Franz Josef, could be around 50 megawatts at peak load.
Cut off, the primary source would likely be local hydropower, along with some diesel plants and distributed generation.
"It would involve using resources already there or specifically committed for this eventuality and trying to do some smart grid based load management.
"We also estimate that there would be some automation facilities that we could integrate into the system, which could be a cheap solution.
"But it also depends on what time a quake occurs, because, if the major source is hydro, there has to be enough of it in that instance - it also depends on how extreme the quake is, and whether the region is really liveable after that."
Nair hopes to come away from the study, funded through the Resilience national science challenge, with a blueprint that could be used in other areas here and overseas following a big disaster.