Benefits
Investment, gas production, royalties, property tax, corporate tax, construction, permanent employment, local procurement and any Canadian research or compute benefit.
Real issues. Real fixes. Real costs.
Data centres have impacts. Many can be reduced. The developer needs to show how, what it will cost and who pays for it.
The debate gets clearer when every issue is put through the same three questions:
If a new industrial load creates a new environmental or infrastructure cost, that cost should be visible in the economics of the project instead of disappearing into a taxpayer, ratepayer, municipality, or future cleanup bill.
Not every project has the same impacts. A dry-cooled gas plant is not a wet-cooled plant. A liquid-cooled data centre is not an air-cooled crypto mine. The mitigation should match the actual design.
| Issue | What matters | Practical mitigation | Who should pay? |
|---|---|---|---|
| CO₂ | New gas-fired generation creates substantial additional emissions. | Highest-efficiency generation, lower-carbon firm power where practical, methane control, carbon capture and permanent geological storage, and durable removal for residual emissions. | Costs attributable to the new load should principally be reflected in the project and the electricity it buys. |
| Water | Total withdrawal and actual consumption must be separated, as must data-centre use from power-generation use. | Dry cooling, closed loops, blowdown recovery, softening/filtration, reverse osmosis, covered storage, reuse and—where justified—near-zero-liquid-discharge systems. | The developer/customer should fund the treatment, storage and supply infrastructure needed because of the project. |
| Noise | Fans, pumps, transformers, generators and heat rejection equipment can create continuous tonal or low-frequency noise. | Setbacks, equipment selection, acoustic enclosures, barriers, operating limits and real measurements at relevant receptors after startup. | The project. Noise mitigation is part of designing an acceptable industrial facility. |
| Power & grid | A hyperscale load can require new generation, substations, transmission and system studies. | Dedicated generation, firm connection agreements, staged load ramps and clearly assigned infrastructure costs. | Project-specific costs should not be quietly transferred to existing ratepayers. |
| Roads & services | Construction traffic and permanent industrial use can impose local servicing costs. | Developer-funded upgrades, road-use agreements, utility servicing and maintenance obligations. | The development creating the incremental demand. |
| End of life | Servers become obsolete quickly; industrial sites and power infrastructure last much longer. | Decommissioning plans, financial security, transferable obligations and periodic updates to the cleanup estimate. | The owner/operator, secured before the liability is left behind. |
Case study: Greenlight
The AUC's June 25, 2026 Greenlight decision provides a newer water estimate than the project's earlier federal filing. The approved project is estimated to require approximately 538,000 cubic metres of water per year. Greenlight uses air-cooled condensers specifically to reduce water requirements.
The AUC describes the source as municipal water supplied by Sturgeon County through a new pipeline loop. The decision does not establish that this supply will be entirely non-potable industrial process water, and it does not itself authorize the water use. Greenlight must obtain the necessary agreements or approvals.
The earlier federal project description estimated approximately 1,250 m³/day. We retain that number as the historical 2025 estimate, while using 538,000 m³/year as the current AUC project estimate.
Case study: Greenlight
Greenlight is a new combined-cycle natural-gas power project being developed in large part to serve Alberta's growing data-centre load. That creates a different policy question than retrofitting an old industrial plant that already exists.
If carbon capture is necessary to make a new source of generation consistent with Alberta's environmental objectives, the cleanest time to price and design it is before construction, not years after the plant begins operating.
Public information is not yet perfectly aligned. Alberta's Major Projects listing says Greenlight will integrate carbon capture and sequestration, while Greenlight's 2025 federal project description described CCS as a potential future addition rather than part of the project as then proposed. That needs a clear, current answer.
A project can use Alberta resources and still be a very good deal. The test is whether the value left in Alberta exceeds the costs left in Alberta.
Investment, gas production, royalties, property tax, corporate tax, construction, permanent employment, local procurement and any Canadian research or compute benefit.
CO₂, methane, water consumption, roads, grid infrastructure, noise, land impact, public incentives, emergency-service demands and eventual cleanup.
Costs caused by the project should be reflected in the project. Shared public infrastructure can still be justified where Alberta receives a genuine broader benefit—but it should be identified openly.
Note: The purpose of this page is to set a test, not to prescribe one engineering solution before final design information exists. Where project documents conflict or evolve, the current binding approval and final design should control.