Is it bad, good, or have we been misinformed?

Common Concerns

The worries people raise about data centres, each one taken seriously and checked. For every concern: what people fear, what the record actually shows, the workaround, the risk that remains even with the workaround, and what a community should demand.

We are not here to calm you down or to scare you. Each concern below gets a plain verdict, then the honest detail behind it. Where the honest answer is “it depends” or “we do not know,” we say so.

Water use

It depends
The worry

That a data centre drains scarce water for cooling, in a semi-arid province, during drought.

What the record shows

There is no single Alberta figure. Cooling designs differ enormously, and most alarming numbers online describe evaporative-cooled facilities on other continents, not the builds proposed here. 111415

The workaround

Closed-loop and direct-to-chip or air cooling recirculate the same water and can cut freshwater use dramatically; some designs use reclaimed, non-potable water.

The risk that remains

“Closed-loop” is a choice, not a guarantee, and it still needs makeup water. Wonder Valley's water licence is being litigated by the Sturgeon Lake Cree Nation, so even a licensed draw can be contested.

What to demand

Make each project disclose its exact cooling design, its water source, and a hard cap, and keep cooling off the potable supply.

The grid & power prices

Real risk
The worry

That new demand strains the grid and pushes up everyone's electricity bills.

What the record shows

The grid can connect just 1,200 MW of new large load through 2028, and it is already fully allocated to two projects, against 16 to 21 GW requested. The grid operator's own CEO says the system cannot connect them all. 1234

The workaround

“Bring your own power”: projects build their own on-site generation so they do not draw on the public grid, and any new generation and transmission can outlast the project.

The risk that remains

On-site power usually means burning gas locally, and grid upgrades can still land on ratepayers if the deals are not structured to make the data centre pay its own way.

What to demand

Ratepayer protection in writing: the project funds its own connection and upgrades, with public transparency on who pays for what.

Emissions & noise

Real risk
The worry

That on-site gas plants add greenhouse gases, air pollution and constant noise next to communities.

What the record shows

Several projects plan to bring their own power, which often means a gas plant. The Alberta Utilities Commission rejected the gas plant behind the Olds project, twice. 38

The workaround

Gas paired with carbon capture, geothermal (Wonder Valley claims gas plus geothermal), or clean grid power where it is available.

The risk that remains

Carbon capture and geothermal at this scale are largely unproven here, and “off-grid gas” moves the emissions on-site rather than removing them.

What to demand

Require a verified emissions and noise plan, with independent monitoring, before approval.

Farmland & land use

Real risk
The worry

That prime agricultural land is paved over for industrial campuses.

What the record shows

Rocky View County rejected a 448-hectare campus proposed by Kineticor Asset Management six votes to one, with residents citing the loss of high-quality farmable soil, and then placed a moratorium on new data-centre proposals. 91213

The workaround

Siting projects on industrial or marginal land rather than prime farmland.

The risk that remains

The best combinations of cheap power, water and flat land often sit right next to farms and towns, so the pressure on farmland is real.

What to demand

Zoning that protects prime farmland, and genuine community consent before rezoning.

First Nations consent

Unresolved
The worry

That projects move ahead on traditional territory without the consultation and consent that treaty rights require.

What the record shows

The Sturgeon Lake Cree Nation is suing the Alberta government over a water licence tied to Wonder Valley that it says was issued without proper consultation, and no provincial environmental impact assessment was required for the project. 1415

The workaround

Real, early duty-to-consult, formal benefit and equity agreements, and First Nations partnership or ownership stakes.

The risk that remains

Consultation can become a box-ticking exercise, and a benefit deal signed later does not fix a licence granted without consent in the first place.

What to demand

Free, prior and informed consultation, honoured treaty rights, and an independent environmental assessment, not a waiver.

Jobs vs. hype

Overblown
The worry

That the promised jobs justify the footprint and the public support.

What the record shows

The construction phase is large, but the permanent workforce is small: Meta and the province cite 3,000+ construction jobs and about 300 permanent, which one analyst framed as roughly $43 million of investment per permanent job. Academics say no independent benefit study exists. 691011

The workaround

Binding local-hire and local-contract commitments, and training programs, written into the agreement.

The risk that remains

Operating jobs are few and specialised, and much of the economic benefit can flow out of the province.

What to demand

An independent benefit study, and local-hire and local-contract commitments in writing before incentives are granted.

Are we being played?

Open question
Our reading, not a source claim
The worry

That opposition to data centres here is amplified while rival economies build them aggressively, so the investment, and the leverage, goes somewhere else.

What the record shows

What is verifiable: the United States and China are racing to build data-centre capacity at enormous scale. What is not verifiable: any coordinated campaign to make Alberta reject them. We have seen no evidence of that, and we will not present a theory as a fact.

The workaround

Judge each project on its own merits, using the cost-and-benefit questions on the rest of this page, rather than on who might gain if we say no.

The risk that remains

The real risk cuts both ways: dismissing every concern as “foreign influence” is just as misleading as accepting every scary number. Both shut down the honest math.

What to demand

Decide on the evidence in front of us. If a project is a genuine asset, prove it on the merits; if it is a genuine risk, fix it on the merits.

Every source is on Citations. See the projects these concerns attach to on the tracker.