All posts
NEM Intelligence· 9 min read

Where to build a data centre in Australia — NEM region by region comparison

Every NEM region offers a different bargain to a large load: cheaper energy, scarcer water, tighter grids or fiercer competition. We compare the five mainland regions across the variables that actually decide a data centre site.

By Zyntax Arc

"Where should we build?" is the most expensive question in digital infrastructure — and the one most often answered with the least data. The National Electricity Market spans five regions, each with its own price behaviour, renewable mix, grid headroom and climate. The right region for a hyperscale campus is rarely the right region for a flexible, renewables-chasing load.

This is a framework for thinking about that decision region by region. It is deliberately directional — the point is not a single ranking, but understanding the trade-offs each region forces.

The variables that actually matter

Before comparing regions, it helps to be clear about what separates a great site from an expensive mistake:

  • Energy economics — wholesale price levels and, crucially, the number of low or negative-price hours a flexible load can exploit.
  • Renewable depth and curtailment — how much clean generation exists, and how often it is curtailed rather than delivered.
  • Grid headroom and MLF — whether the network can accept a large new load, and the loss factor it will carry.
  • Water and climate — cooling burden and water availability, increasingly a hard constraint.
  • Competitive density — how crowded the region already is with existing data centres.

A region can be brilliant on one axis and disqualifying on another. The job is to weigh them for your load.

Region by region

NEM regionEnergy economicsRenewable depth & curtailmentGrid & MLFWater & climateDC competition
NSW1 (New South Wales)Higher average prices; deep, liquid demandGrowing, REZ-led; moderate curtailmentCongested corridors; MLF varies sharply by nodeTemperate; water generally workableVery high — the established hub
VIC1 (Victoria)Moderate prices; solid liquidityStrong wind; rising solarMature network, pockets of constraintCool climate aids coolingHigh
QLD1 (Queensland)Volatile prices; strong solar middaysAbundant solar; real curtailment riskLong network; MLF discipline essentialWarmer; cooling burden higherModerate and rising
SA1 (South Australia)Frequent low and negative-price hoursVery high renewable share; curtailment commonSmaller grid; firming often neededHot, dry; water is a key constraintLower — room to differentiate
TAS1 (Tasmania)Hydro-stable pricingHydro + wind; low carbonConstrained interconnection (Basslink)Cool, wet; cooling-favourableLow

Directional summary. Actual outcomes are node-specific and change with every AEMO update.

How to read the trade-offs

A few patterns fall out of the table.

If you can flex, chase the negative hours

A load that can shift or modulate — or that co-locates with storage — is rewarded most by regions with many low and negative-price hours. South Australia and, at times, Queensland lead here, precisely because high renewable penetration drives prices down (and negative) when the sun and wind are strong. The catch is curtailment and firming: the cheap energy is only valuable if you can actually take it, which often means on-site batteries.

If you need firm, always-on capacity, weigh stability

A latency-critical, must-run facility values predictability. Tasmania's hydro-anchored pricing and cool climate are attractive, but interconnection limits can cap how much new load the region absorbs. Victoria offers a balance of mature grid, decent renewables and a cooling-friendly climate.

If you need scale and ecosystem, accept the premium

New South Wales remains the gravity well — the deepest demand, the most connectivity, the largest existing cluster. You pay for it in higher energy prices, congested corridors and fierce competition for sites. For some operators the ecosystem is worth the premium; for others it is exactly why differentiated sites elsewhere win.

Everywhere, the node beats the region

The single most important caveat: regional averages hide enormous variation. Two connection points in the same region can differ by millions in delivered cost once MLF, curtailment and water are applied. Choosing a region narrows the search; it does not make the decision.

The region sets the weather. The connection point decides whether you have a good business.

From framework to defensible decision

This kind of comparison is a starting point — useful for orientation, dangerous as a final answer. The reason most teams stop here is that going deeper has historically meant weeks of specialist analysis per site.

That is the gap Zyntax Arc is built to close. DC Intelligence scores any specific site across eight layers — energy economics, MLF, curtailment, NABERS readiness, government alignment, water, competitive density and satellite land assessment — and rolls them into a single, defensible verdict with a 10-page report. Flex Load Intel does the complementary job at NEM scale, ranking 1,296 assets for flexible-load and co-location potential.

The result is the ability to move from "South Australia looks interesting" to "this connection point in this region scores 82/100, here is exactly why, and here is what would change the answer" — in minutes, not months.

The bottom line

There is no universally best NEM region for a data centre. There is only the best region for your load, and within it, the best connection point you can defend with data:

  • Flexible, storage-paired loads are rewarded by renewable-heavy regions with negative-price hours — if they solve curtailment and firming.
  • Firm, always-on loads favour stability and cooling-friendly climates — within grid limits.
  • Scale-and-ecosystem plays accept a premium for depth and connectivity.

Pick the region for the trade-off it forces. Then let the data pick the site.

Keep reading

All posts
Energy Economics· 7 min read

Understanding MLF and why it costs data centres millions

Marginal Loss Factors are one of the most overlooked numbers in Australian energy economics — and one of the most expensive. Here's how MLF works, why it moves, and how a single decimal can erase millions from a data centre's business case.

Read more

Ready to make better infrastructure decisions?

Join the teams using Zyntax Arc to score sites, model NEM economics and pressure-test compliance before they commit capital.