Alberta Just Approved a Project That Will Emit About 100 Million Tonnes of Greenhouse Gases In Its Lifetime. Here's What We Could Build Instead.

Lori Guetre
Jul 28
6 min read

Updated: 3 days ago

And why today’s choices are important for the next 37 projects in the queue.

A new natural gas power plant is being built in Sturgeon County to power Meta's first Canadian data centre. It's called the Greenlight Electricity Centre (GLEC) - 932 megawatts, $4.6 billion, developed by Pembina Pipeline, Morgan Stanley, and Kineticor. It's the first of 19 more gigawatts of data centre projects lining up to build in Alberta. Real jobs, real capital, anchor infrastructure for the AI economy Canada wants to build.

Big projects are exciting. Alberta has world-class wind. Its natural gas is among the cheapest in North America. It hosts the world's largest-capacity CO₂ pipeline, running through the same municipality where the new plant is being built - infrastructure that could transport captured CO₂ to safe, permanent geological storage. Alberta gets more sunshine than most of Germany, which is the world’s leader in installed solar. Every ingredient for clean industrial leadership is right there.

So why does the plant being built today look like this?

Line drawing of an industrial smokestack labelled "Current GLEC Configuration," with red text above reading "Legally allowed to emit roughly 100 million tonnes of CO2e in its lifetime."

Legally allowed to emit roughly 100 million tonnes of carbon dioxide equivalent (CO₂e) in greenhouse gas emissions over 30 years. From one project. That's about 15% of Canada's annual emissions, cumulatively, from a single facility - the equivalent of adding about 700,000 gasoline cars to Canadian roads for the next thirty years.

I've spent the last two weeks pulling this thread. The full analysis is in a paper I published Monday: Seven Ways to Power a Data Centre. What I found surprised me.

Why the plant looks like this

In November 2025, Canada suspended the federal Clean Electricity Regulations (CER) for Alberta. Since then, no federal emissions performance standard applies to new gas plants in Alberta - only Alberta's own Technology Innovation and Emissions Reduction (TIER) regulation, which lets a facility be fully compliant while doing very little to actually reduce its emissions.

Data centres need 24/7 firm power, and this one is asking for almost a gigawatt (GW). The major constraint facing the GLEC project designer is that the Alberta Electric System Operator (AESO) grid is not ready to take on these megaprojects. The only way to go ahead with a megaproject in Alberta right now is to “bring your own power” – essentially “islanding” any new power source from the grid.

A natural gas (methane) plant is a reliable way to get firm power, and natural gas is abundant in Alberta. Renewables like wind and solar are cheaper per megawatt hour (MWh) than fossil fuels in Alberta, but hybrid configurations have more complexity and risk to manage. And without the grid for stability and reliability, an islanded hybrid configuration would require the project to become its own mini grid operator and would need to overbuild gas and battery storage to handle renewable intermittency.

Venn diagram showing four overlapping circles labelled "Move Quickly," "Need a Lot of 24/7 Firm Power," "No Grid Connection Available," and "Essentially Free to Pollute" - the last shown in red. A small illustration of a polluting factory sits at the centre intersection of all four.

Facing the constraints and conditions listed above, the rational choice for the GLEC plant developer is to build an unabated natural gas plant. It's also, by our calculations, the most expensive path to Canada's legislated 2050 net zero commitment.

Here's what the numbers actually look like. Seven different ways to power the same data centre, using only Alberta resources, under four different emissions frameworks. We also look at two types of configurations: “islanded” (today’s constraint) and “grid connected” (the more normal situation for new projects in Alberta and across the world).

Bar chart comparing seven plant configurations under four emissions frameworks, with the leftmost bar (current Alberta TIER standard) showing the highest total cost at geological net zero, and the rightmost bar (grid-connected geological net zero) showing the lowest. Icons below each bar show the technology mix - gas with carbon capture, wind and solar with battery storage, low-emissions methane, permanent carbon dioxide removal, and grid interconnection. A red row at the bottom shows lifetime emissions in millions of tonnes of CO2e - 104 for the current plant, dropping to zero for the geological net zero configurations.

If you read the chart left to right, you’ll see that the leftmost bar (the plant being built) has the lowest base cost of the islanded configurations today (blue) but the biggest green section on top, representing the eventual cost of removing 100 million tonnes from the atmosphere. The rightmost bar (Configuration 4G, grid-connected with new renewables) costs 20% less for the raw electricity and 60% less than the current plant once we account for reaching net zero. Same power. Same Alberta resources. Dramatically different outcome.

You’ll also see the different technologies that naturally get pulled in by a developer solving for “lowest cost compliant” as emissions requirements tighten. With any kind of emissions constraint, it makes sense to pull in renewables - even under islanded conditions. The developer would also want to implement carbon capture and storage (CCS) because avoiding those emissions is cheaper than pulling them back out of the atmosphere afterward. Lower-emissions methane further reduces the total emissions from the project (red row at the bottom), and permanent carbon dioxide removal (CDR) removes the small amount of residual emissions left over to get the project all the way to geological net zero – where we need to get to in order to halt global warming.

Two surprises, one credit card

Surprise 1: Building the plant as approved is like putting it on a credit card.

Today’s purchase price (leftmost blue bar) looks low - about $81 per megawatt-hour. But Canada has legislated net zero by 2050, and the interest is accumulating in the atmosphere the whole time. Someone eventually pays. The cleanest islanded build (Configuration 4) costs more up front but arrives at net zero for 30% less, because at-source abatement is dramatically cheaper than removing emissions from the atmosphere afterward.

The bill comes due in 2050. Pay a bit more up front, or pay much more later. That's the choice.

Surprise 2: Clean electricity is cheaper than unabated gas - when the plant can connect to the grid.

This is Texas's story. Texas built the enabling infrastructure - the "plumbing" for new megaprojects - like fast grid interconnection, mature power purchase agreement (PPA) markets, and streamlined renewables permitting. Now its data centres are being paired with hundreds of megawatts of new solar at speed, without any state mandate to do so. Alberta has all the natural advantages Texas has and some Texas doesn't.

Build that plumbing in Alberta and clean electricity costs 20% less than the plant currently being built on delivered energy alone - and 60% less once we account for reaching net zero.

When grid connection is available, the “essentially free to pollute” condition is no longer needed – the other three conditions can be satisfied without it. Here is what the same picture looks like when we fix the grid connection problem.

Revised Venn diagram showing the same four circles as before, but now with a green checkmark on "Grid Connection Available" (replacing "No Grid Connection Available") and a red circle-with-slash over "Essentially Free to Pollute." The centre intersection now reads "Texas-style hybrids."

Projects can rapidly add cleaner energy sources and want to do so because they are lower cost. AESO handles firming at grid scale with lower-emissions natural gas plants that are either built with or retrofitted with CCS. The small amount of residual emissions can be removed by permanent CDR.

The choice is being made now

None of this is a critique of AESO or GLEC’s developers – they are solving rationally for the framework they’ve been given. The question is whether we can give them a better framework.

These new megaprojects are important opportunities for jobs and economic development, and it is difficult for electricity infrastructure to move quickly. But it is not necessary to let this project put another 100 million tonne burden on future generations at higher cost, and there are 37 more projects in the queue asking for over 19 more GW of power.

Three windows are open right now that will shape how these 37 projects get built:

Alberta's AI data centre policy framework, being finalized this month.
AESO Phase II Large Load Integration, being designed in parallel.
Canada's Sustainable Finance Taxonomy consultation, closing August 13, 2026.

If Alberta modernizes grid interconnection as Texas has done, Configuration 4G becomes the template. The Alberta Carbon Grid gets anchor demand. New-build wind and solar gets 15-20 GW of hyperscaler PPA volume this decade. Alberta becomes a leader in CO₂ sequestration. Canadian permanent carbon removal capacity grows.

Every subsequent project gets easier and cheaper.

Bar chart showing a single blue bar for Configuration 4G at approximately $65 per megawatt-hour, with technology icons below indicating wind, solar, and battery storage; permanent carbon dioxide removal; and grid interconnection. Lifetime emissions shown as zero.

Alberta has everything it needs. The wind is better than Texas. The gas is cheaper. The CO₂ pipeline is already running through Sturgeon County. The geology is here. The engineering is here. The capital is here.

What's missing is the plumbing. And asking our governments to build it.

What you can do

Share this information. Important choices are being made now that are not aligned with Canada’s net zero by 2050 commitment. The projects being built today will last for decades and future generations will be left with the burden and the cost.

If you know someone at Alberta Energy, Environment, AESO, or one of the affected federal ministries – reach out. The plumbing can potentially be fixed faster with more resources and political will behind it.

We will be submitting comments to the federal Taxonomy consultation and we will make them public. Share your comments or simply re-share ours to add weight.

Possible by Design has a Canadian federal e-petition open right now - e-7510, on mandatory climate footprint labels. It's the first of four advocacy ideas that would help close the systemic gaps we describe in our Agency Architecture and create the transparency we need. Please sign it, and please share it.

The other three advocacy ideas are ready and waiting. Sign this one, then come see the full set - because the hard part is behind us. What's left is building the social permission for smart policy. That part is on all of us.

Get this right, and Canada and Alberta avoid adding 100 million tonnes of CO₂e to the atmosphere from this one project alone. And we build the template for the next fifty years of industrial infrastructure - on our own resources, on our own terms, with our own workforce.

The choice is being made now.

Read the full analysis: Seven Ways to Power a Data Centre | Companion workbook