How to choose a data centre site in Australia: the five tests that decide viability.

Every operator entering Australia, and every developer pivoting into data centres, faces the same expensive question. Is this site genuinely viable, or does it just look viable. The difference is worth tens of millions of dollars and one to two years of schedule. A site can pass on price, on location and on first impressions, and still fail on the one thing that actually stops the build. The way to avoid that is a disciplined test applied before capital is committed.

At DivineLab Worx we validate every candidate site against five pillars. A site has to pass all five to be recommended. Here is what each one actually checks.

1. Power proximity

Power is first because it fails the most projects. The tests are specific: distance to the nearest zone substation, available headroom in megavolt amperes, whether a dedicated substation is required and who bears that cost, transmission proximity for loads above fifty megawatts, and whether a dual feed is possible for the redundancy tier the client needs. In 2026 this pillar has become decisive, with AEMO reporting about 5.4 gigawatts of data centre demand in the connection queue and an approximate two year path from application to energisation. A site with no substation headroom and no funded upgrade pathway is not a data centre site, whatever else it offers.

2. Zoning clearance

The second pillar asks whether you can actually get planning consent, and how long it will take. That means the current zone classification and permitted uses, overlay constraints such as flood, bushfire, biodiversity and heritage, proximity to sensitive receptors like homes and schools, the council's demonstrated attitude to data centre applications, and whether the project triggers a state significant or major project pathway. A heritage listing or a one in one hundred year flood zone can end a project before it starts. Better to know on day five than after settlement.

3. Fibre path

Connectivity is the third test. A serious facility needs route diverse fibre, not a single provider. The checks are distance to the nearest tier one exchange or point of presence, dark fibre availability from more than one carrier, latency to the major exchanges, and the feasibility and cost of new duct if diversity is missing. New fibre builds can take twelve to thirty six months and run to tens of thousands of dollars per kilometre, so a site that looks cheap on land can be expensive on connectivity.

4. Water access

Water has moved from a technical footnote to a public and regulatory flashpoint. In 2026 an OpenAI backed Australian facility dropped a recycled water cooling plan, drawing scrutiny, and industry voices are pressing developers to manage water and energy responsibly rather than free ride on shared resources. The pillar tests proximity to a water main of sufficient capacity, the water authority's willingness to supply, recycled water availability for cooling, and wastewater discharge rights. Getting this wrong is not only a cost problem. It is a social licence problem that can stall an approval.

5. Regulatory and work health and safety readiness

The final pillar covers the conditions that turn a site into a lawful, insurable, operable asset. Contamination status for former industrial land, asbestos risk on existing structures, flora and fauna obligations, construction environmental management, building code compliance, and, increasingly, the security obligations that attach once a facility is treated as critical infrastructure. This is where the SOCI Act, fire standards and environmental assessment all live. Clients consistently underestimate the time these take, which is exactly why they belong in the validation, not the aftermath.

Why a structured test beats a property search

Commercial property agents will find you land. They generally cannot confirm that power is actually available, that the fibre is diverse, or that the water authority will supply. Large engineering firms can do the technical work, but slowly and at high cost. The gap in the middle is a fast, technically grounded validation that tells a decision maker, in weeks not quarters, whether to proceed, to proceed on defined conditions, or to walk away. That is what compresses the time to build ready, and time to build ready is what actually costs money in this market.

How DivineLab Worx delivers it

Our critical infrastructure and deployment engagements run candidate sites through all five pillars and return a client ready site intelligence report: an executive summary, a pillar by pillar assessment, a ranked risk register, a timeline to build ready and a clear recommendation. We work across the critical infrastructure sector for operators entering Australia, developers expanding a footprint, and investors testing an opportunity before they back it. If you are weighing a site right now, the cheapest thing you will ever do is find out early whether it passes.