Comparing Canadian Provinces’ Approaches to Building Data Centers

In this article:

• How Alberta, Saskatchewan, Manitoba, British Columbia, Ontario, and Quebec are approaching AI data centre development.

• What makes each province attractive to data centre developers and investors.

• How increasing electricity demand from AI data centres could affect residents.

• Which provinces are best positioned for future data centre growth.

The latest center of attention in the media, and in public discourse, is the rapidly growing number of data centers in North America. It also seems that the more this technology is discussed, the further it spreads, covering every acre of farmland and placing pressure on natural resources. However, what not everyone realizes is that data centers have been around since the 1940s. Almost a century ago they were called mainframe computers, but the core of the concept was the same: although computers got smaller, data centers expanded in order to power today’s digital services such as Google and ChatGPT requests.

At its core, data centers are establishments containing rows of racks with hundreds of computers stacked on top of one another. Because those computers need to be continuously supplied with power, data centers use both on-site power distributors and backup generators. Additionally, data centers are equipped with cooling systems that ensure stable temperatures and humidity levels inside the building.

Artificial intelligence is not the only field that requires large-scale data centers. Governments, universities, and some large corporations often use in-house data centers that store servers and other IT equipment. There are also retail data centers that offer server racks at a price-per-kilowatt fee. Overall, out of more than 11,000 data centers worldwide, Canada houses about 250.

AI data centers, on the other hand, are resource-intensive and complex infrastructures that, besides requiring major land holdings, also use substantial amounts of electricity and water, and operate under arrangements involving multiple levels of government, local stakeholders, Indigenous communities and utility providers. 

It comes as no surprise that more Canadians are raising concerns about proposed data centers in their communities. Such projects often involve financial and environmental trade-offs that developers do not always fully disclose. However, to understand the impact of AI data centers in Canada, we must look at how each province approaches this evolving technology.

Factors to consider:

• Structures: 

The right corporate structure for a data center is based on whether the facility includes on-site power generation or primarily focuses on data processing and storage. A data center that locates power generation with computing operations requires a different legal profile from a facility that operates solely as a storage facility.

• Financing: 

AI data centers are high-capital projects, and financial agreements cover a framework which aligns capital resources with available incentive programs, risk management and project stages. While most AI data center projects are financed through a blended capital structure and/or corporate debt, in other cases, the government offers loan guarantee programs. Of course, the way a data center project will get financed depends on long-term revenue certainty, risk profile, and the developer’s commitment and overall strategy.

• Energy supply:  

Depending on the province’s regulations and the proposed data center project, this raises the question of how a facility will secure energy supply and maintain grid reliability under sustained demand.

Provinces Overview

Alberta

Alberta operates a competitive wholesale market, where the Alberta Electric System Operator (AESO) manages the power grid and supports the wholesale market. In this deregulated electricity market, generators and consumers have an opportunity to negotiate private power purchase agreements, while data center developers are allowed to purchase power from independent generators of their choice. Alberta is moving towards ensuring that developers will “bring their own generation” in order to limit their reliance on the Alberta grid. 

Compared to Ontario and British Columbia, Alberta also attracts data center investors and developers with abundant industrial land availability and an overall business-friendly environment. Moreover, building in a province that is a leader in wind and solar development, developers have a chance to pair renewable energy resources with natural gas. 

However, based on AESO reports, recent data center connection requests exceed approximately 16,000 MW (this reflects requested load, not guaranteed demand, and is significantly higher than current peak demand of roughly 12,000 MW). To navigate increasing demand, AESO introduced a process that would limit the volume of large-load connections in the near future. Data center projects that require grid connection would have to wait for AESO development of a long-term integration framework.

• Projects in development and grid connection requests: According to Alberta’s project map, there are currently 31 operating, under-construction, and proposed data centre projects across the province.

The AESO has approximately 40 large data center and AI projects under review that could collectively add significant electricity demand if all proceed.

• Impact on residents:

Expansion of large AI data centers in Alberta raises questions about substantial electricity demand and whether residential electricity bills would be impacted in the long term. There are two scenarios that could affect Alberta’s residents and the electricity system: the first is where developers fund grid upgrades and required generation to run their facilities, which would have a limited impact on consumers, and the second is where new data centers could potentially generate tax revenue and local employment, contributing to a broader balance between grid pressure and economic growth. 

Saskatchewan

The Saskatchewan Power Corporation (SaskPower), which is owned by the provincial government, is the primary electricity provider and regulator of electricity across Saskatchewan. In contrast to Alberta’s deregulated wholesale electricity market, in Saskatchewan, AI data center developers cannot negotiate PPAs with independent power generators. 

To secure power supply in Saskatchewan, large-scale data centers are required to work directly with SaskPower through phases of transmission interconnection and coordination across SaskPower, SaskTel and SaskEnergy, which support the facility.

Despite a much smaller grid (compared to Ontario, Alberta or Quebec), which limits how many AI data centers could be supported, Saskatchewan offers relatively stable electricity pricing and a less congested grid that is largely powered by natural gas, coal, and wind energy. 

• Projects in development and grid connection requests:

Saskatchewan currently has one major confirmed large AI data center project in development (Bell Canada’s AI facility near Regina). However, unlike Alberta or Ontario, Saskatchewan does not yet have a publicly visible pipeline of grid-connected data center requests.

• Impact on residents:

Since Saskatchewan’s data center growth is not as rapid as Alberta’s, the impact on residents is also limited. However, as electricity demand grows, and new hyperscale facilities require additional power and transmission infrastructure, there is potential for upward pressure on system costs. Whether this affects residential electricity bills will depend on how developers are required to fund infrastructure and how costs are allocated. Saskatchewan is at a pivotal point where the government is seeking to balance affordable electricity for residents with economic diversification through large-scale industrial development.

British Columbia

Unlike the deregulated Alberta market, large-scale data centers in British Columbia face challenges in regulatory access. According to BC Hydro policy frameworks, the process of accessing electricity has shifted from administrative interconnection to strategic allocation and planning decisions. 

Recently proposed and developed large-scale projects must compete for electricity allocation, potentially requiring developers to pay higher rates or accept supply constraints during peak demand. This reflects an infrastructure-first approach, combining phased development with the province’s hydroelectric system.

Despite these regulatory constraints, BC’s electricity system is primarily powered by renewable hydroelectric resources, making it one of the cleanest grids in North America, which makes the province particularly attractive to hyperscale companies with net-zero commitments.

• Projects in development and grid connection requests:

While BC’s established data center market is concentrated in the Lower Mainland, new hyperscale AI projects remain limited compared to Alberta. Instead of approving all large-scale developments, the province focuses on managing future growth and existing operators, while requiring AI data center developers to compete for access to provincial electricity capacity.

According to BC Government data, BC Hydro reports approximately 6,800 MW of industrial connection requests. This is more than six times the generating capacity of the Site C hydroelectric project.

Such demand has prompted the province to move away from a first-come, first-served model for AI and data centres under its Electricity Allocation Framework. The province has reserved 100 MW for data centres and 300 MW for AI projects, based on expected economic and community benefits.

• Impact on residents:

Growing demand for electricity has prompted BC to adopt a more selective approach in order to protect residential ratepayers and maintain grid reliability. While data centers are expected to contribute to innovation and employment, the framework also ensures that they do not compromise affordability or reliability for households and businesses.

Manitoba

Manitoba’s approach to AI data centers is a case-by-case evaluation framework that prioritizes smaller computing facilities over hyperscale and energy-intensive projects. The province defines its approval framework based on emissions constraints, regulatory requirements, and project size. 

Manitoba Hydro, as the sole electricity provider in the province, requires developers to go through an evaluation system where requests for large electrical loads are defined as either high, medium or low priority.

Additionally, every AI data center developer must comply with Manitoba Hydro requirements related to sustainable building practices and energy efficiency reporting.

• Projects in development and grid connection requests:

According to Statistics Canada, there are currently eight data centers in Manitoba; however, none of them are hyperscale facilities.

So far, the province has been rejecting proposals for large AI data centers due to high electricity demand and infrastructure constraints.

• Impact on residents:

Although data centers currently have limited impact on Manitoba residents (as they remain relatively small and are supported by renewable electricity), over time, increased AI-driven demand may increase pressure on electricity systems. While large-scale data centers would require additional investment in generation and transmission infrastructure, the impact on electricity rates will depend on how Manitoba Hydro manages system expansion and cost allocation.

Ontario

Ontario is considered Canada’s largest and most mature data centre market. The province’s digital infrastructure and connectivity make it attractive to data center developers.

However, rapid AI-driven demand also creates challenges around electricity supply. The IESO operates Ontario’s electricity market, while the Ontario Energy Board (OEB) oversees regulatory oversight and rate structures. Data centers are required to register as energy market participants and comply with system planning and connection requirements. Data centers are also expected to participate in regional planning processes that may affect provincial electricity systems.

• Projects in development and grid connection requests:

There are over 80 data center facilities in Ontario. The province expects data center demand to become a major driver of future electricity growth.

More than 6,500 MW of data centre load has requested connection to Ontario’s electricity grid. This is equivalent to approximately 30% of Ontario’s 2024 peak electricity demand, or roughly the output of the Bruce Nuclear Generating Station.

Forecasts suggest that over the next ten years there will be at least sixteen more data centers connected to the grid, representing approximately thirteen percent of projected electricity demand growth.

• Impact on residents:

As Ontario’s data center market continues to grow rapidly, concerns are emerging about future transmission capacity. In response, the province has introduced a new approval process that prioritizes projects expected to strengthen the economy and create jobs. Although increased infrastructure investment may impact utility bills over time, the government emphasizes that electricity should be allocated based on broader economic value rather than automatic approvals.

Quebec

Quebec is characterized by abundant low-cost hydroelectricity and a cold climate, making it one of Canada’s leading locations for hyperscale and AI data centres.

Because Quebec follows a sustainable development model (powered by Hydro-Québec) and prioritizes renewable hydroelectricity, data centres are also built around high-density computing workloads.

On top of that, Quebec has implemented frameworks which ensure that new data center proposals align with grid capacity and provincial energy planning.

• Projects in development and grid connection requests:

Existing operators and hyperscale facilities (primarily in Montreal) are expected to see electricity consumption increase sevenfold by 2035, exceeding 1,000 MW. In response to this anticipated surge, Hydro-Québec has proposed a new electricity rate for large data centres of approximately 13¢/kWh, and continues to require all projects over 5 MW to undergo a competitive selection process.

Rather than approving every connection request, Quebec now prioritizes projects that deliver the greatest economic benefits and make the most strategic use of the province’s renewable electricity resources.

• Impact on residents:

The expansion of AI data centres in Quebec can generate investment, employment, and tax revenue while strengthening the province’s position as a digital hub in Canada. However, their intense electricity requirements create competition for Quebec’s clean hydroelectric resources. To protect long-term energy affordability for residents and ensure grid reliability, Hydro-Québec has adopted a more selective allocation process for large data centres.

Frequently Asked Questions:

Which Canadian province is best positioned for future data centre growth? 

While there is no definitive answer as to which province has a clear advantage in the AI data centre industry, Alberta appears best positioned for rapid expansion. This is primarily due to Alberta’s flexible electricity market, which attracts data centre developers and investors. However, Ontario’s strengths lie in market demand and digital connectivity, British Columbia offers sustainability credentials, and Quebec provides clean, renewable-powered AI infrastructure. Overall, provinces are expected to continue balancing electricity availability, regulation, and long-term infrastructure planning.

Do data centres create long-term economic benefits? 

Although this remains a debated issue, data centres can potentially deliver economic benefits through increased tax revenue, capital investment, and employment in digital infrastructure. However, high electricity demand and relatively limited permanent employment mean these benefits depend on whether projects meaningfully contribute to broader economic growth and infrastructure development.

Will AI data centres change how provinces plan their electricity systems? 

Yes. The large and continuous electricity demands from data centres are driving new investment in transmission and distribution infrastructure while prompting utilities to reconsider grid capacity abilities. As requests for new data centres increase, provinces are incorporating this demand into long-term planning while also evaluating whether new projects deliver sufficient economic value. As a result, electricity planning is shifting from simply meeting demand to strategically allocating resources based on economic, environmental, and infrastructure-wide priorities.

Key takeaways:

• AI data centres are reshaping provincial energy planning across Canada.

• Electricity demand is driving policy changes with large AI data centres requiring continuous, high-volume power and long-term infrastructure investment.

• While data centres can generate investment, tax revenue, and employment, provinces have to balance those benefits against electricity costs and infrastructure requirements.