ANALYSIS: There seems a strong chance a $700m plan to build a massive data centre just north of Invercargill will succeed.
But it probably pays to be realistic about the number of jobs that could bring to the region, which will need to find a variety of ways to plug the gap left by the eventual closure of the Tiwai Point aluminium smelter.
Back in 2014, I gave Datagrid founders Remi Galasso and Malcolm Dick a 70 per cent chance of successfully building the $445m Hawaiki trans-Pacific internet, which they eventually did in 2018.
Datagrid is in many respects a simpler proposition, whose prospects come down quite literally to cold and relatively predictable economics.
* Internet cable connecting NZ, Chile and Australia could create data centre industry in Southland
* $445m NZ internet cable opens for traffic
* Now we have a broadband motorway, it’s time to put the foot down
Demand for cloud computing capacity has been rising steadily around the world, bringing locations for data centres that wouldn’t have appealed five or 10 years ago into play.
Operators are looking for five things; cheap and ideally renewable power, proximity to large numbers of computer users, reliable internet connectivity, political stability, and ideally cheap land to build on.
A cool climate cuts power bills and can make the overall equation add up.
In warmer climates a large proportion of the electricity that data centres consume will be needed to cool rather than just to power the dense racks of computer servers they house.
With an all-around average temperature of just under 10 degrees, Southland’s climate is about as good as it gets.
Data centres are best built reasonably close to big populations to avoid the drawback of lag when computer users access the services and data they house.
But if 35 milliseconds of latency is acceptable to the big cloud computing firms, then Southland ticks the box to service the bulk of the Australian population.
What has held back the development of a data centre industry in Southland up to now has been the lack of diverse international connectivity options.
But Datagrid’s plan to lay two new subsea cables to Invercargill should solve that. Having Hawaiki Cable as a sister company obviously helps.
Once again, Galasso and Dick may have got their timing right.
New Zealand’s seismic risk shouldn’t be a showstopper.
Datagrid would be a “tier 3” data centre on a five-point scale denoting the back-ups it would have in place and the tolerances for outages, putting it in the mainstream of the mass market.
“The whole world is looking for safe havens to have redundant copies of their data so even if it’s not a primary location a lot of businesses will be using it as a back up,” Dick says.
“With most of the big data centres, wherever you put your data it is also simultaneously being replicated in two other places in the world.”
Data centre capacity is a commodity and if Datagrid can deliver powered rack space at a cheap enough price, then it should get customers and financiers – simple as that.
Galasso and Dick have proved through Hawaiki Cable and their other past ventures that they are ‘doers’ rather than ‘talkers’, and there is no real reason to think they are likely to have got their sums wrong.
Data centres are intentionally designed to be places where people are rarely needed.
The equipment they house is as far as possible managed and configured remotely.
So they are not big employers.
A lot of workers would be required during construction, but Galasso estimates Datagrid would employ just 25 staff once completed.
Could its mere presence encourage IT companies and start-ups to spawn nearby?
It’s a nice thought, but I’m not sure quite why it would.
In the US, local governments often provide incentives to attract large data centres, but evidence concerning the scale of the employment benefits they bring appears to justify some scepticism.
Queenstown in nearby Otago appears to be coming into favour with more start-ups for reasons that are more to do with its lifestyle.
But I’d imagine the extra skilled jobs that the data centre itself would bring to the region might be in the dozens rather than the hundreds.
The new cable connectivity might have more spin-offs, particularly if it does also link up the Chatham Islands, Stewart Island and Antarctica.
But, even then, I wouldn’t expect Datagrid to provide tremendous ‘bang for the megawatt’ jobs-wise when compared to the Tiwai Point smelter.
The same might be said for a plan to make the region a centre for the production of ‘green hydrogen’, which took a step forward on Friday when Meridian and Contact chipped in $2m for a feasibility study.
Meridian’s generation manager Guy Waipara says that project could be quite “elastic” in terms of its scale.
If they hope to replace the smelter’s jobs, Southland’s civic leaders will need to also attract more manufacturing industries to the region with some of Manapouri’s freed-up power.
There are options to do that I think.
From an environmental perspective, Datagrid would have to be a good thing.
Like the smelter, it would effectively be exporting renewable power, but in its case in the form of ‘bits and bytes’ rather than aluminium ingots.
Instead of displacing coal and gas-fired aluminium smelters in China and the Middle East, Datagrid would be most capping the demand for additional fossil-fuelled data centres in Sydney and Melbourne.
That won’t show up in New Zealand’s carbon accounts, but would be good for the planet.
That’s just so long perhaps as its racks aren’t used – as unfortunately many are in Iceland – for the unproductive environmental disgrace that is Bitcoin mining.