Demand is real but overstated: contract conversion lags years behind announcements, and ~6 of 7 requested megawatts are phantom
Pipeline totals range from 6 GW to 44 GW depending on what is counted, and roughly six in seven requested megawatts never arrive. The four screens that turn connection requests into a planning band, and why scarcity and glut coexist.
Two correction lenses before any demand figure is used. First, the conversion lag: NEXTDC's contracted 740 MW (Jul 2026) sits against 110.9 MW billing at FY25: the sector's revenue is a J-curve trailing its headlines by years. Second, phantom demand: AEMO received 44 GW of connection requests against ~6 GW required under Step Change; Oxford Economics estimates ~6 of every 7 requested MW will never materialise.
AEMO connection requests vs need, GW
| Item | Required (~6 GW) | Phantom (~38 GW of 44 GW never materialises) |
|---|---|---|
| Connection requests to AEMO (2025 IASR, national) AEMO 2025 IASR | 6 | 38 |
| Required to meet AEMO Step Change demand AEMO 2025 IASR | 6 |
NEXTDC: contracted vs billing, MW
| Item | Value |
|---|---|
| Billing at FY25 (110.9 MW) | 110.9 |
| Contracted at FY25, Jun 2025 (244.8 MW) | 244.8 |
| Forward order book, Jul 2026 (565 MW) | 565 |
| Contracted utilisation, Jul 2026 (740 MW) | 740 |
Why both can be true
Genuine demand can be at record highs (68% pre-commitment, take-up exceeding supply) while headline pipelines are simultaneously inflated 7x. The discipline is to underwrite contracted MW with power secured, and to treat everything else as optionality. Optionality is not supply.
Deliverability check: only ~25% of the total pipeline is 'currently fully deliverable' given power, water, cost, lead-time and regulatory barriers (M3/Colliers, Nov 2025).
Pipeline totals range from 6 GW to 44 GW depending on what you count: only ~25% is fully deliverable
Six ways to count the Australian pipeline
| Pipeline measure | Figure | What it counts |
|---|---|---|
| AEMO connection requests (2025 IASR, national) | 44 GW | Every application, incl. speculative; Oxford Economics: ~6 of 7 MW phantom |
| Knight Frank pipeline (2025) Knight Frank (Apr 2025) | ~15 GW | UNVERIFIED (poor OCR; register item 3) |
| NSW project pipeline (Climate Council/DC Byte) Climate Council (Jun 2026) | 11.4 GW / 44 projects | Announced projects, all stages |
| M3 built-out trajectory (early 2030s) | ~9.5 GW | Delivered capacity incl. committed pipeline |
| AEMO Step Change requirement (2030) | ~6 GW | What system demand actually needs |
| Fully deliverable near-term (M3/Colliers) | ~25% of pipeline | Power, water, cost, lead-time, regulatory screens |
How inashanu discounts the pipeline
- 1. Power screen Does the project hold a connection agreement, or only an enquiry? (AirTrunk SYD3 had 300 MW committed against a 400 MW project at one point.)
- 2. Contract screen Binding lease or MoU? (OpenAI-S7 is an MoU; CDC's 555 MW is binding.)
- 3. Funding screen Debt/equity committed? (AirTrunk's ~A$4.3bn Kemps Creek facility was still being sought in Jun 2026.)
- 4. Planning screen SSD approval vs application vs SEARS.
Applying these screens turns 44 GW of requests into the 5-8 GW planning band (inashanu synthesis). Section 10 takes its low end as the bear case; base and bull are built bottom-up.
inashanu deliverability rule of thumb
Count a pipeline MW at 100% if contracted + powered + funded; 50% if two of three; 10% if one; 0% if connection-enquiry stage. Read qualitatively across the Section 05 table it gives ~2.5-3.5 GW of high-confidence additions to 2028, consistent with CBRE's 1.8 GW live-capacity path plus committed UC. Run arithmetically: 1,350 MW operating, plus 68% of the 2,301 MW under construction, plus 25% of the 5,892 MW planned, reaches 4,388 MW by 2030 (Market_Sizing B26). The 1,612 MW gap to the 6,000 MW base case (B27) is what that case asks of projects not yet visible.
CBRE's shortfall paradox
CBRE's 2026 outlook projects a ~1.5 GW shortfall for Australia despite tier-1 APAC status: scarcity and glut coexist because the constraint is deliverability (power + planning) and not the volume of announcements.
Sources
- ITK Research, as at May 2026
- Shiny Side Out, as at Jul 2026
- AEMO 2025 IASR
- Oxford Economics
- Kalkine, as at 21 Jul 2026
- M3 Property via The Industrialist, as at 24 Nov 2025
- Knight Frank, as at Apr 2025
- Climate Council, as at Jun 2026
- M3 Property, as at Nov 2025
- de Valence, as at Dec 2025
- Certified Strategic/CBRE, as at 25 May 2026
- Troview, as at Jun 2026
Related
- What is in Australia's data centre development pipeline?
- How much data centre capacity will Australia have in 2030?
- Does Australia have 1.35 GW or 2 GW of data centre capacity?
- How much electricity do Australian data centres use?
The full primer is published as an 85-page PDF (download) with the underlying market model as a workbook (download).
This page is one of 30 in Australian data centres: the research, indexed by the question each one answers, with the four pages where the published sources disagree set first.
The research is published by inashanu, a studio that produces fixed-scope market research for Australian private equity, priced and dated before work starts. The sample engagement is this primer in full.
This standard, on your question. Six fixed-price deliverables, A$14,000 to A$45,000 excluding GST, opening at the A$14,000 market map in 10 business days. State the decision and the date it has to be made; the written assessment of whether the evidence base will carry it follows within two business days and carries no charge. Request a brief assessment · the catalogue and its prices.