To:
Hon Simeon Brown
Hon Winston Peters
Hon Shane Jones
Hon David Seymour
Hon Brooke van Velden
Hon Chris Bishop
Hon Nicola Willis
Hon Simon Watts
Hon Judith Collins
31 March 2026
Subject: Datagrid NZ’s AI Factory: A Strategic Misallocation of Critical Electricity Resources
“New Zealand tends to focus on short-term gains at the expense of long-term outcomes.”
Sir Bill English
The current fuel situation has made one point unmistakably clear:
New Zealand’s reliance on global energy systems is no longer tenable.
When the media is reduced to counting inbound shipments to determine whether diesel supply will hold, it is evident that the assumptions underpinning our energy security have already failed.
In that context, your attention is called to the proposed Datagrid New Zealand AI factory development in Southland, being advanced by a local council ill-equipped to assess national interest.
As outlined, this project would become New Zealand’s second-largest electricity consumer, after the Tiwai Point aluminium smelter. Its initial demand of 280 MW, potentially rising to 1 GW, places it in a category of national strategic impact, not merely a commercial development as currently assessed.
In plain English, a single 1 GW data centre consumes roughly the same amount of electricity as would be required to electrify New Zealand’s entire heavy transport diesel fleet.
Presented as a $3 billion investment, a substantial proportion will be directed offshore to imported equipment, with the local component largely confined to construction. Although the facility will generate rates revenue for the host council, this must be weighed against the national cost of allocating scarce renewable electricity to an activity with limited ongoing economic integration.
In effect, a localised fiscal incentive is being allowed to shape the use of a critical national strategic asset, with consequences that extend well beyond the interests of the host community.
The issue is whether allocating somewhere in the range of 5–17% of New Zealand’s renewable electricity to a globally-oriented AI processing operation serves the long-term interests of this country, at a time of growing vulnerability in liquid fuel supply, particularly diesel, which underpins the transport, agriculture, and the physical economy. In this context, electricity is now New Zealand’s most strategic domestic energy asset.
Parliament’s response should be direct and unambiguous for any data centre, not just this one:
- Declaration: Electricity, water and their infrastructure are critical sovereign strategic assets
- Self Sufficient: Data centres are welcome, provided they are fully energy self-sufficient
- Self-Funded: Data-export operators must fund, build and operate new generation capacity equal to 100% of their projected consumption, additional to the national grid
- Grid Neutral: They must not draw on existing grid capacity
- Water Neutral: They may use but must not deplete local water resources for cooling
- Export: If exporting computation, operators must fund energy and water use, not Kiwis
Sovereign Security Consent: Subject to strict compliance with the standards set out above, qualifying developments should follow a national rapid-approval pathway, with consent granted by Cabinet on the advice of the Minister following national consultation, as a matter of energy sovereignty and control of a critical national asset, while not tying up operators in red tape.
National Priority, Strategic Direction and Energy Sovereignty:
This page of the brief links approval of the Datagrid AI Factory to the unfolding liquid-fuel crisis.
It is not about saving the planet or Green virtue signalling; it is about preserving first-world capacity.
It is not about welcoming data centres; it is about electricity as the future lifeblood of the economy.
The 20th century was the era of cheap oil. That era is over. NZ’s future lies in electricity. Unless NZ elects to build nuclear power plants, it will require all the clean electricity it can generate.
The Datagrid proposal presumes there is surplus electricity, even as it would become one of the largest users in the country, and potentially the largest at full scale. That presumption is false. NZ does not have surplus renewable electricity. Additional demand increases reliance on imported dirty fuels. At the end of long and fragile supply chains, that is a direct risk to national security.
In the hierarchy of energy, the lessons of the past month should be clear, NZ must move to energy sovereignty that lies with electricity not thermal combustion. In short, starting right now, NZ must adopt, as a matter of national priority, a shift away from diesel-powered heavy vehicles to electric.
In the attached appendix, data centre load versus diesel displacement is analysed. The numbers are sobering. One 8.8 TWh/year data centre approximately represents the displacement value of New Zealand’s heavy equipment (trucks, tractors, buses and ferries) diesel consumption. In other words, if Datagrid operates at its proposed maximum (1 GW continuous), it would consume the electricity that otherwise could be used to displace diesel. In that context, history will record the diversion of NZ’s electricity to offshore computation as one of the greatest strategic errors this century.
Displacement will not happen overnight. Some of NZ’s fleet can be converted today; the rest as technology advances. China, which is diesel-import dependent, is aggressively reducing diesel dependence with heavy investment in battery technology and battery-swapping charging stations to keep freight moving. NZ can be a first-adopter; but not if we give away the electricity needed to power the heavy EVs. We are ideally placed to become a scalable national platform for heavy EV.
There are two ways to initiate the shift to energy sovereignty (diesel displacement):
- Centralised national procurement: Analogous to Pharmac, Central government buys heavy EVs from overseas in volume, negotiating national-level pricing; buying vehicles easy to maintain locally. Prioritise durable designs, interchangeable parts and quick-swap-out batteries over planned obsolescence. Limit the proliferation of makes and models to enable standardisation. Once the heavy EVs arrive in NZ, utilise existing wholesale and retail distribution networks. Or, alternatively:
- Domestic production through partnership: Partner with an overseas manufacturer to establish a heavy EV manufacturing facility in NZ (perhaps in Bluff, South Island), producing trucks, tractors, buses and ferries, particularly those designed for NZ conditions such as logging trucks and automated farming systems. Invest in robotics to reduce the impact of higher labour costs, and institute national battery swap-out charging regime. Pursue this as a matter of national security rather than purely commercial production economics, but make it profitable for the private sector.
The question is simple: does the Government commit another strategic error, this time by ceding electricity to offshore computation, or does it use that electricity to secure economic sovereignty?
At a time when global systems are demonstrably fragile, reserving that energy to replace imported fuels and strengthen the domestic economy warrants direct ministerial and Cabinet intervention, rather than a business-as-usual decision driven by local consenting processes.
Regardless of the feasibility of moving ICE to EV, approval of the AI factory should, as a matter of national security, be conditional on 100% self-generation of its projected energy demand.
Annex: Energy Allocation – Data Centre Load vs Diesel Displacement
All figures are indicative and intended to show order-of-magnitude comparisons.
1. Scale of Proposed Electricity Demand
The proposed Datagrid NZ facility has:
- Initial load: ~280 MW
- Potential expansion: up to ~1,000 MW (1 GW)
Annual electricity consumption:
- 280 MW (continuous):
≈ 2.5 TWh per year - 1 GW (continuous):
≈ 8.8 TWh per year
For context:
- Total New Zealand electricity generation is approximately 40–45 TWh per year
- The Tiwai Point aluminium smelter is one of the country’s largest single users of electricity, consuming on the order of 5 TWh per year
At full scale, the Datagrid facility would be comparable to, or larger than, the largest existing single industrial electricity user in New Zealand and would represent a substantial share of national electricity demand.
2. Diesel Use in the New Zealand Economy
New Zealand consumes approximately:
- Diesel: ~3–4 billion litres per year
Diesel is a primary energy source for:
- Freight transport (road and rail)
- Agriculture and contracting
- Construction and heavy machinery
Energy content:
- Approximately 10 kWh per litre total annual diesel energy:
- On the order of 30–40 TWh (thermal)
Diesel imports represent a multi-billion-dollar annual cost to the New Zealand economy and are critical to the functioning of the physical economy.
3. Electrification Efficiency Advantage
Electric systems are significantly more efficient than combustion-based systems:
- Internal combustion drivetrain efficiency: ~30–40%
- Electric drivetrain efficiency: ~80–90%
On this basis, replacing diesel use with electric systems would require:
- On the order of 10–15 TWh of electrical energy, depending on system efficiencies and application mix
4. What 280MW–1GW Represents in Real Terms
- 280 MW (≈2.5 TWh/year):
→ Represents a meaningful share of the electricity required for partial diesel displacement - 1 GW (≈8.8 TWh/year):
→ Represents a substantial proportion of the electricity required to materially displace diesel use across key sectors
At full scale, the proposed data centre load is of a similar order of magnitude to the electricity required to significantly reduce New Zealand’s dependence on diesel.
5. Realistic National Fast-Charge or Swap Network
For a true battery-swap system, the physical swap itself is likely to be in the 3 to 5 minute range once standardized heavy-truck systems exist at scale.
- With ranges of 500–600 km, swap stations should be spaced at about 350 km intervals
- This aligns with driver rest requirements (30 min. every 5.5 hours – ~75 km/h ≈ 400 km)
- Requirement: Standardised swappable battery architecture
- Initial deployment should prioritise long-haul corridors (e.g. SH1)
6. Strategic Trade-Off
The allocation decision is therefore not marginal.
It is a choice between:
Option A – Data Centre Allocation
- Consumes several terawatt-hours per year
- Comparable in scale to the country’s largest industrial electricity users
- Serves primarily offshore computational demand
- Does not reduce fuel import dependence
Option B – Domestic Electrification
- Uses equivalent electricity to displace diesel imports
- Strengthens energy security
- Reduces exposure to volatile global fuel markets
- Supports transport, agriculture, and industrial continuity
7. Policy Implication
Every terawatt-hour allocated to offshore-facing data processing is a terawatt-hour not available for replacing imported fuels in the domestic economy.
At the scale proposed, this becomes a national allocation decision, not a commercial one.
Conclusion
The proposed Datagrid facility should be evaluated not only as an investment, but as a competing claim on New Zealand’s most strategic energy resource.
- The question is not whether the project can be powered.
- It is whether this is the highest-value use of that power for the New Zealand economy.