Alberta, It's Time to Build

On July 2, 2026, the Prime Minister and the Premier of Alberta stood together in Calgary and announced a new pipeline to the West Coast. One million barrels a day, Bruderheim to a deepwater port in southern British Columbia, permissions targeted so construction can start as early as September 2027.

Six days later, Meta announced a $13 billion AI data center in Sturgeon County. Its first in Canada. Its largest anywhere outside the United States. One gigawatt of power, fed by a new plant that's permitted to nearly double that. To put a gigawatt in perspective, Alberta's entire grid peaked at about 12,200 megawatts in 2024, so this single campus is roughly 8 percent of the province's record demand.

For a hundred years, Alberta's business model was simple: pull energy out of the ground, ship it somewhere else, let someone else turn it into something valuable. The pipelines are the biggest version of that old model.

The data center is the first serious version of a new one, where the energy stays home and gets converted into the most valuable product in the world right now, compute. Central Alberta is going to run both models at once. Nobody else in Canada can do that at scale, and I think the next ten to fifteen years will be defined by what compounds out of it.

The old economic model

Start with the pipeline, because it funds everything else.

Egress has been the binding constraint on Alberta's economy for two decades. Anyone who managed money through 2018 remembers what happens when production grows faster than egress (pipelines and rail): the differential between what Alberta crude sells for and what world markets pay blows out, and billions in royalties and free cash flow evaporate. Trans Mountain's expansion in 2024 relieved some of that pressure. The new line adds another million barrels a day, roughly a 20% increase in Canada's total crude export capacity, and it points at Asia instead of Cushing, Oklahoma.

The stated ambition attached to this project is to roughly double provincial oil production toward eight million barrels a day over the next decade. Will it get all the way there? Maybe not. The direction matters more than the destination. Filling a million new barrels of pipe requires a full development cycle, meaning in-situ projects, processing, steam, power, diluent logistics, camps, contractors. Ottawa and Alberta figure the pipeline plus the Pathways carbon capture project will create something like ~175,000 jobs across the buildout. The last time Alberta ran a cycle like that, roughly 2005 through 2014, it rebuilt the demographics of the whole province. We got younger. Our universities boomed. Wages ran hot for a decade. I believe that same cycle is queued up again.

But if that were the whole story, you'd be right to shrug. Another boom, another bust. Alberta has seen this movie.

The new economic model

The Meta announcement matters more than the pipeline, even at a third of the size.

Let's work out why.

Think about what a data center does. It takes raw energy, natural gas in our case, converts it to electricity, then converts the electricity into intelligence that gets sold globally at software margins. Definitionally, that's a token refinery. It just refines BTUs into tokens instead of gasoline. Alberta has spent a century exporting the feedstock and importing the finished product. For the first time, the full cycle runs inside the province. Gas produced here, burned here, consumed by servers here, generating value that gets reinvested here.

People have talked about diversifying Alberta's economy since before I was born, and it almost always meant diversifying away from energy. Tech corridors. Film credits. Whatever Ontario was doing with its real estate industry... Most of it failed, because you can't out-compete your own comparative advantage. This buildout is different. It's diversification downstream of energy rather than away from it. Our edge, cheap and abundant natural gas, happens to be exactly the input the fastest-growing industry on earth needs most.

The way Meta structured the deal tells you it scales, too. The facility is anchored by the Greenlight Electricity Centre, a new gas-fired combined cycle plant in Sturgeon County built by Pembina, Morgan Stanley Infrastructure Partners, and Kineticor. Phase one is 932 megawatts at a cost of roughly $4.6 billion. It's permitted to double to 1,864 megawatts, built in modular blocks of about 466 megawatts each, and contracted to the data center under a long-term tolling agreement. The whole campus sits under Alberta's bring-your-own-power rules, so Meta pays the full cost of its generation and the operational infrastructure around it. Sturgeon County's announcement actually says the arrangement will cut costs for Alberta ratepayers by up to 6 percent. The hyperscalers are building their own power plants, and the neighbours' bills go down.

The gas math

This section is for anyone who wants to model what all this does to gas demand. The numbers are public, and the arithmetic is friendly. I've pulled much of this from the press releases and Semianalysis datacenter models.

A gigawatt of data center running around the clock consumes 24,000 megawatt-hours per day. A modern combined cycle plant burns roughly 7 gigajoules of natural gas per megawatt-hour. Multiply those together, convert the units, and one gigawatt of gas-fired data center load needs about 160 million cubic feet of gas per day. Call it 60 bcf a year.

You don't have to take the napkin's word for it either. Pembina's own disclosures say Greenlight's first phase, at 932 megawatts, will increase gas demand in the Industrial Heartland by roughly 150 to 160 mmcf per day, rising to about 320 mmcf per day at the full 1,864 megawatt buildout.

So here's the metric worth memorizing: every gigawatt of gas-fired data center capacity in Alberta means a bit more than 0.16 bcf per day of new, in-basin, baseload gas demand.

Now scale it.

The Greenlight site at its full permitted 1,864 megawatts is about 0.3 bcf/day, per Pembina. If the thesis of this essay is right and one campus becomes five, then five campuses at Meta's opening scale of a gigawatt each is roughly 0.8 bcf a day of incremental demand inside Alberta. If their power plants build out the way Greenlight is permitted to, the number heads toward 1.5. For context, LNG Canada's first phase pulls about 2 bcf a day of feedgas, and getting there took two decades of politics plus a pipeline over the Rockies. The data center path delivers roughly half an LNG Canada of demand growth with no coastal pipe, no tankers, no export permit, sold to investment-grade counterparties who consume every hour of every day.

And the quality of this demand matters as much as the quantity, maybe even more. Gas demand has always been seasonal and weather-driven, which is why AECO prices spend half the year embarrassing the producers. Datacenter load has no seasons. It's flat. It's contracted. It's creditworthy. It supports the basin's price floor in exactly the months that currently hurt most. Greenlight has already locked up transportation through the Alliance Heartland expansion and the NGTL system, which tells you the midstreamers see the same thing I do. For Montney and Deep Basin drillers, this is a structurally better customer than the one they've been waiting for LNG to become, and it showed up in their own backyard.

Why one becomes five

Hyperscalers copy each other on the infrastructure side. Their site checklist is pretty consistent: land, power, water, fiber, cool climate, sane regulator, workforce.

Sturgeon County and Alberta just passed a de facto audit from $META. It isn't a reach to think $MSFT, $ORCL, $AMZN and $GOOG are watching and running Alberta through their own models right now. The second data center is easier than the first. The transmission gets planned. The county learns how to permit. The trades get trained. The gas turbine supply chains, the substation contractors, the fiber routes, the water agreements... all of it exists now. Greenlight reached final investment decision in July 2026 with service expected in the second half of 2030. That's roughly four years from decision to power. In most American markets, the interconnection queue alone runs longer than that. Alberta's grid operator already has a long line of proposed data center load behind META. Some of it will never get built. Enough of it will.

The Phoenix precedent

If you want to understand what happens downstream of an anchor investment like this, go research what happened after TSMC picked north Phoenix.

TSMC's first Arizona commitment was $12 billion in 2020. It grew to $65 billion. Then $165 billion, the largest foreign direct investment in American history. Then in July 2026 the company added another $100 billion, taking the total to $265 billion.

The direct numbers are enormous. But the knock-on economic benefits are the real story.

Fourteen of TSMC's key suppliers committed to building their own US facilities, several in Arizona, covering industrial gases, chemicals, and equipment components. Since 2020, Arizona has attracted more than 40 semiconductor projects worth over $102 billion, and Greater Phoenix now hosts more than 75 semiconductor companies. The Greater Phoenix Economic Council projected the first three fabs alone would generate more than 20,000 construction jobs, over 18,000 indirect supplier and consumer jobs, and $1.4 billion in state tax revenue over 13 years. There's a 2,300-acre master-planned development called Halo Vista going up beside the campus, a $7 billion project with housing, labs, offices, and retail, planned around roughly 20,000 residential units. Even Semicon West, the industry's flagship conference, left San Francisco for Phoenix after more than fifty years.

None of those suppliers or developers or conference organizers were in the original announcement. They came because the anchor made the region investable. That's the pattern central Alberta can expect if we get this right. Meta is the anchor. The suppliers, contractors, and developers who orbit hyperscale construction are the follow-on, and they make their location decisions two to four years behind the anchor. Which means the window to position for them is open right now, while everyone else is still arguing about whether the boom is real.

The flywheel

Big anchor loads change the economics of energy for everyone else. This is the part I keep coming back to.

A gigawatt of guaranteed, creditworthy, 24/7 demand is the best customer a power developer can have. It underwrites new generation that would never get financed on merchant risk alone. New generation at scale pushes the cost curve down, and because the anchor tenant pays its own freight, ratepayer costs fall rather than rise as the buildout proceeds. Multiply that by five campuses and central Alberta ends up with one of the deepest, cheapest, most reliable industrial power markets on the continent.

Cheap reliable power is the oldest industrial recruiting tool there is. It's why aluminum smelters sit next to hydro dams. Petrochemicals follow the same logic. So do hydrogen, fertilizer, metals processing, greenhouse agriculture, and eventually manufacturing. Each new industrial tenant deepens the labour pool and the supplier base, which lowers costs for the next tenant. Demand funds generation. Generation lowers cost. Low cost attracts industry, industry funds services and talent, and the services and talent make the next project cheaper and faster to build. Around it goes.

Who actually builds this

Every one of these projects is, at bottom, a labour and contractor story. I started my career on the tools side of Alberta's industrial economy, before I ever managed a dollar of anyone's money, so this section is personal. It's worth being specific about who does the work.

1. Project 1 — Greenlight. The day it reached final investment decision, Aecon announced a $1.2 billion EPC contract to build it. Underneath that contract sits a whole stack of services: civil earthworks, deep foundations, structural steel, turbine rigging and heavy haul, high-pressure pipefitting, instrumentation, and thousands of hours of electrical work terminating into a new substation.
2. Project 2 — The data center campus. Electrical work is the single largest cost line in data center construction, routinely a third or more of the build, followed by mechanical and cooling. Meta's general contractor is Frost Collective, a joint venture between Clark Builders and PCL. Two firms with deep Edmonton roots. That tells you how much of the prime contracting value stays local.
3. Project 3 — The pipeline. Right-of-way clearing, stringing and welding crews, directional drilling under every river crossing, pump stations, and a marine terminal.

Now add up what central Alberta will be running simultaneously in the early 2030s. A million-barrel pipeline. One of the world's largest carbon capture networks. A multi-phase power plant. At least one gigawatt-class data center, plus all the housing and municipal infrastructure that follows. The trades those projects need overlap almost completely: electricians, powerline technicians, welders, pipefitters, millwrights, power engineers, crane operators, instrumentation techs. Three megaprojects bidding for the same journeymen. Wages in these trades are going to do what oilfield wages did in 2006.

Alberta also holds an advantage here that never shows up in a site selection spreadsheet. We already have the industrial construction ecosystem. The module fabrication yards in Nisku and Fort Saskatchewan that assembled oilsands process units can assemble power plant modules and data hall skids. The camp operators, the crane and rigging outfits, the inspection firms, the electrical contractors who wired the upgraders... it all transfers. Most regions courting data centers have to import that ecosystem from scratch. We built ours over the last 100 years and then watched it run below capacity for over a decade.

The knock-on economy

Humans are bad at thinking exponentially. We model the first-order effects and miss the forest for the trees. The construction stack is the first wave. The second wave is everything that services the construction stack, and it's broader than most people realize.

Start with real estate services, because land moves first in every cycle like this one. Industrial brokers who know the Heartland corridor are about to have the best decade of their careers. So is everyone behind them: appraisers valuing land with no recent comparables, surveyors and geomatics firms staking sites, title and conveyancing work, environmental firms clearing Phase Ones and Twos, civil engineering shops designing the subdivisions and stormwater systems and road connections that turn farmland into serviced industrial parks.

Then the development layer. Land assembly specialists. Industrial developers building spec warehouses for suppliers who need space on eighteen months' notice. Residential builders meeting worker demand, property managers, and the hospitality operators who build the hotels and extended-stay properties that every megaproject region runs short of. Phoenix is instructive here too, because the market response to TSMC wasn't just houses. It was an entire master-planned district built to hold the ecosystem.

Then professional services, which scale with the complexity of the projects rather than their tonnage. Law firms first: regulatory and permitting work, Indigenous consultation and equity agreements, procurement and EPC contracts, land acquisition, water rights, municipal planning files. Accounting and tax practices handling project entities, cost segregation, SR&ED claims, and the personal planning needs of newly wealthy contractors (I see this one up close). Engineering and environmental consultancies at every stage from feasibility to commissioning. Insurance brokers and surety bond providers, because a $1.2 billion EPC contract doesn't get signed without bonding capacity behind it. Banks and private lenders financing contractor working capital, equipment, and developer land loans. Recruiting firms, immigration lawyers, payroll providers moving and managing thousands of workers. Even IT and physical security firms, since hyperscalers impose their vendor standards on everyone who touches their sites.

Notice something about both lists. These are mostly local businesses. The turbines come from abroad, and the chips come from Taiwan and South Korea, but the lawyer, the appraiser, the electrician, the broker, and the accountant are in Edmonton, St. Albert, Sherwood Park, and Fort Saskatchewan. That's how a megaproject's spending actually reaches a regional economy.

The wealth effect, and who pays the taxes

Now follow the money into the households, because this is the part that decides whether the boom stays popular enough to last. Wealth effects in a region like this work through three channels.

1. Wages come first. Trades and professional incomes rise as three megaprojects compete for the same people, and higher incomes flow straight into local spending, savings, and small business formation.
2. Asset prices come second. Homes, industrial land, and local businesses all get repriced as demand arrives, and every existing owner gets wealthier without lifting a finger.
3. The third channel is the tax base, and it's the one nobody talks about. When billions of dollars of industrial assessment lands in a municipality, the cost of running that municipality gets spread across a much larger base, and the share carried by each household falls.

Alberta already has a forty-year demonstration of this running next door to the new boom. Shell's Scotford complex, the refinery, chemicals plants, and upgrader near Fort Saskatchewan, sits in Strathcona County alongside the Refinery Row industrial strip. That industrial assessment base carries a huge share of the county's budget, which is why Strathcona County residents enjoy some of the most favourable property taxes of any urban area in Canada. When CBC compared Sherwood Park with Saint John, New Brunswick, a city with a remarkably similar mix of homes, retail, and oil refining, it found similar houses, coffee shops, and shopping centres paying 30 to 45 percent less tax in Strathcona County, while the refineries there paid several times more than Saint John's refinery. Same services. Better amenities. Lower household bills. The industry pays because the industry is there, and because Alberta's municipal system actually lets counties tax heavy industrial property properly.

Now hold that up against Edmonton. City council approved property tax increases of 8.9 percent for 2024, 6.1 percent for 2025, and 6.9 percent for 2026. That's a cumulative increase of roughly 29 percent in four years, and city figures show the average household paying about $816 per $100,000 of assessed value in 2026. I don't think those increases are purely the product of waste. They're what happens when a growing city's costs are carried almost entirely by homes and small commercial properties, with no heavy industrial base to share the load. The burden sits on families because there's nothing else to put it on.

Which is what makes the Sturgeon County story so consequential. A $13 billion data center and a $4.6 billion power plant are about to become taxable industrial assessments in a county of roughly 20,000 people, on top of Meta's $60 million commitment to local roads and water, with the project expected to generate about $250 million a year for the province in royalties, taxes, levies, and fees. Sturgeon County is on a path to becoming the next Strathcona County, a place where the corporation carries the municipal budget and the family's tax bill flattens or falls. Every subsequent data center, gas plant, and pipeline facility deepens that base.

The policy lesson runs both directions. For the region's municipalities, revenue-generating infrastructure is the only sustainable alternative to taxing households harder, so the competition to host it, and the regional agreements to share it, deserve as much council attention as any service file. For Edmonton specifically, the lesson is sharper. The city can't annex its way to a Scotford. It can compete for the data centers, logistics facilities, and power infrastructure that fit inside its boundaries, and it can negotiate regional cost-sharing that reflects where the workforce actually lives. A city facing 29 percent tax growth in four years should treat industrial assessment as the tax relief program it is. Every gigawatt that lands in the region transfers municipal burden from the family to the corporation. That, more than any jobs headline, is how ordinary households experience an infrastructure boom going right.

Abundance, tested in the field

Ezra Klein and Derek Thompson's book Abundance makes a simple argument: the defining problems of the Western world are supply problems. Not enough housing. Not enough energy. Not enough transmission. And the shortage is self-inflicted. Over fifty years, rich democracies layered on procedures, reviews, and veto points until the systems designed to prevent bad projects became systems that prevent projects, full stop. Their evidence is brutal and specific. California authorized high-speed rail in 2008 and still hasn't connected two cities. San Francisco takes years to approve a single apartment building. Clean energy projects across America die not from lack of capital but from interconnection queues and litigation that outlast the developers. The book's core claim is that scarcity is a policy choice, and the political movement that learns to build again, at speed, will own the next era.

Central Alberta over the next fifteen years is the closest thing North America has to a live test of that claim. Greenlight went from announcement to final investment decision to a shovel-ready $4.6 billion power plant with a four-year build. Meta committed $13 billion within roughly a year of the province opening the door. The west coast pipeline went from provincial pitch to a federal Major Projects Office referral with a construction target eighteen months out. Those timelines are not normal on this continent, and that gap is our opportunity. Alberta isn't out-competing Virginia and Oregon on capital access. It's out-competing them on the willingness to say yes quickly. In an era where compute demand doubles faster than any regulator's review cycle, speed of approval has become a resource in its own right.

The other half of the abundance argument applies here too, and we shouldn't kid ourselves about it. Klein and Thompson's villains usually aren't opponents of growth. They're supporters of growth who attach so many conditions, carve-outs, and process requirements to each project that the project quietly dies. Alberta isn't immune. The moment approvals slow, transmission lags, or housing supply gets rationed by the same zoning reflexes that ground down Vancouver and Toronto, the advantage evaporates, because the hyperscalers' checklist has no loyalty in it. Abundance here isn't a slogan. It's an operating discipline, measured in one number: how many months from application to token. Guard that number, improve on it, and the Alberta boom compounds.

Talent is the real bottleneck

Meta's construction phase alone peaks around 3,000 workers, with more than 300 permanent operations roles after that. The pipeline and Pathways buildouts need tens of thousands more. And the people needed are not, for the most part, software engineers.

Alberta already has the best trades culture in Canada. The arithmetic above says demand will exceed it anyway. NAIT, SAIT, and the polytechnics need to be funded and told, explicitly, to double throughput in electrical and power trades. The U of A should be building the country's best programs in power systems, grid engineering, and data center operations, because those graduates will have five employers within an hour of campus. Immigration policy, both federal nominee allocations and provincial credential recognition, becomes a competitive tool. The region that solves trades supply can win the marginal project.

Real estate and city building

Edmonton is one of the last major cities in North America where a young family can buy a house on ordinary incomes. That affordability is now a strategic asset in the talent war, and the fastest way to destroy it is to let a demand shock hit a supply-constrained housing market. Every city planner in the region should be studying what data center booms did to housing in northern Virginia and Columbus, and what the TSMC buildout did to north Phoenix, and doing the opposite of the mistakes.

The specific geography matters here. Sturgeon County, Fort Saskatchewan, Bruderheim, Gibbons, Morinville, and northeast Edmonton sit at the center of both catalysts. Industrial land in that corridor is going to reprice. It probably already is. Serviced lots, workforce housing, and commercial nodes along Highway 28 and the Heartland corridor are the plays. So is the boring work: water infrastructure, road upgrades, and the annexation and zoning decisions that determine whether growth happens in an organized way or a resentful one. Meta's $60 million for local roads and water should be treated as the opening precedent.

Where the capital goes, and what the policy has to do

For investors, the exposure ladders neatly. Midstream and power infrastructure sit at the center of both catalysts, and it's no accident that Pembina shows up in the pipeline consortium, owns nearly half of Greenlight, and supplies its gas through Alliance. Natural gas producers with Montney and Deep Basin acreage get a new in-province demand source, flat and contracted, layered on top of LNG. Power generators and transmission owners get a decade of contracted growth. Industrial land and logistics real estate in the Heartland corridor gets repriced. And the services layer, everything from engineering firms to equipment rental to modular builders to the professional firms in the knock-on economy, is where the highest-torque, least-crowded opportunities live.

The pattern from every previous cycle holds. The commodity gets all the bad headlines. The infrastructure gets the early returns. The services get the multiples on their earnings.

Policy decides whether this is a five-year boom or a thirty-year platform, and the list isn't complicated. Permit power generation and transmission on timelines measured in months, not years, and publish those timelines so the world can see them. Let large loads bring their own generation, and make the Meta model, where the anchor tenant pays its full freight, the standard, so residential ratepayers never become a political casualty of this boom. Pre-zone and pre-service industrial land at provincial scale. Fund trades and professional education. Lock in Indigenous equity participation early, because the pipeline's timeline will depend on it. Get ahead of water allocation in the North Saskatchewan basin before it becomes a constraint. Structure regional tax and revenue sharing so the whole metro benefits from industrial assessment, not just the county that holds the land. And keep the tax and royalty framework boring and stable for as long as possible, because nothing kills a capital cycle faster than bad economics and a government that changes the rules mid-game.

Alberta has been given a strange gift. The last great cycle of the old energy economy and the first great cycle of the new one, arriving in the same decade. We've been looking for a way to diversify this economy my entire adult life, and the way finally showed up twenty-five minutes outside Edmonton.

Data centers can be the transition.

The region that builds this well gets rising wages, rising home equity, and a tax bill carried by corporations instead of families.

Alberta, it's time to build.

Sources

Pipeline and Pathways:

• Canada and Alberta advance west coast pipeline project proposal and Pathways Project — Prime Minister of Canada, news release, July 2, 2026. One million bbl/d capacity, MPO referral, Indigenous equity, ~175,000 combined jobs.
• Strengthening our sovereignty, diversifying our exports, reducing emissions, and building a stronger economy — PMO speech transcript, July 2, 2026. Pathways as the world's largest carbon capture project, 16 Mt annual emissions reductions, Building Canada Act listing targeted by October 1, 2026.
• West Coast Oil Pipeline — Government of Alberta project page. Submission to the Major Projects Office July 2, 2026, partnership with Trans Mountain Corporation and Pembina Pipeline, national interest listing expected by October 1, 2026.
• Government-owned Trans Mountain to lead $35-44 billion oil pipeline to BC Lower Mainland — Canada's National Observer, July 2, 2026. Route following the Trans Mountain corridor to southern BC, cost range, construction targeted as early as September 2027.
• Agreement reached on multibillion-dollar Pathways carbon capture and storage project — CBC News, July 13, 2026. Pathways MOU signed as the condition for the pipeline; tanker port in southern British Columbia.

Meta Sturgeon Data Centre:

• Hello, Sturgeon County! — Meta Data Centers blog, July 2026. 1 GW campus, partner list including Clark Builders, PCL, Greenlight LP, AltaLink, Capital Power, and the AESO.
• Breaking Ground on Meta's First Data Center in Canada — Meta newsroom, July 2026. CA$13B+ investment, 33rd data center globally, ~3,000 peak construction workers, 300+ operational jobs, ~CA$60M in local roads and water infrastructure, closed-loop dry cooling, full cost of energy use paid by Meta.
• Meta building its first Canadian data centre northeast of Edmonton — CBC News, July 2026. Premier Smith's statement that the project will generate at least $250 million annually for Alberta; $60M local infrastructure commitment.
• Meta to build $13B data centre north of Edmonton, its first in Canada — ConstructConnect / Journal of Commerce, July 2026. Frost Collective joint venture between PCL Construction and Clark Builders as delivery partner.
• Meta plans gigawatt-scale data center in Alberta, Canada — Data Center Dynamics, July 2026. Meta fully funding new generation and grid infrastructure, improving reliability for all consumers.

Greenlight Electricity Centre:

• Canada's Pembina Advances Gas-Fired Data Center Project — EnergyNow / Bloomberg, July 2026. FID on the 932 MW, C$4.6B Greenlight Electricity Centre in Sturgeon County; completion scheduled for the second half of 2030.
• Meta's Canadian AI Data Center: A New Model for Infrastructure and Energy Integration — Data Center Frontier, July 2026. Ownership split (Pembina 47.5% / MSIP 47.5% / Kineticor 5%), Aecon and Técnicas Reunidas fixed-price EPC, Siemens turbines, ~85% of costs de-risked, combined cycle design.
• Pembina to Secure Data Center Deal with Meta in Alberta — Zacks, 2025. Greenlight designed as an ~1,800 MW station built in phased modular increments, with gas demand disclosures of ~160 mmcf/d at phase one rising to ~320 mmcf/d at full buildout per the Pembina/Kineticor October 2025 project update.

TSMC Arizona precedent:

• TSMC Arizona — TSMC official page. Investment expanded from $12 billion in 2020 to $265 billion, the largest foreign direct greenfield investment in American history; July 2026 announcement of additional fabs.
• TSMC Deepens U.S. Chip Manufacturing Push With $100B Arizona Expansion — citybiz, July 2026. The additional $100 billion bringing the total to $265 billion across 12 facilities; supplier network establishing a self-contained ecosystem in Greater Phoenix.
• TSMC Arizona Anchors $7 Billion 'City Within a City' as Phoenix Emerges as Silicon Desert — Signal, July 2025. Supplier relocations to Phoenix since TSMC's announcement, Halo Vista's 2,300 acres and industrial/research buildout.
• $7 billion development Halo Vista breaks ground near TSMC — KTAR, April 2026. Halo Vista groundbreaking; comparison to Taiwan's Hsinchu Science Park.
• Transforming the Desert Edge: Projected Growth around TSMC's Phoenix Campus — In Business Phoenix, October 2025. Investment history from $12B, fab-by-fab expansion, ~6,000 high-skilled manufacturing jobs.

Property taxes and the wealth effect:

• Edmonton-area homes, stores, malls pay lower property taxes than Saint John, but oil refineries pay more — CBC News.
• Edmonton city council approves 8.9% property tax increase in 2024 — Global News, April 2024.
• Edmonton city council sets property tax increase at 6.1% for 2025 — CBC News, December 2024.
• Edmonton city council approves 6.9% property tax increase for 2026 — CBC News, December 2025. Includes the ~$816 per $100,000 of assessed value figure for 2026.
• Property taxes — Strathcona County official tax pages.

Other:

• AESO 2024 Annual Market Statistics — Alberta Electric System Operator. 2024 seasonal peak records of 12,384 MW (January 11) and 12,221 MW (July 22); total installed capacity of 23,122 MW.
• Ezra Klein and Derek Thompson, Abundance, Avid Reader Press, 2025.
• Gas conversion arithmetic (author's calculation): 1 GW x 24 hours = 24,000 MWh/day; ~7 GJ/MWh combined cycle heat rate; ~1.05 GJ per mcf; approximately 160 mmcf/d per GW, cross-checked against the disclosed 150 to 160 mmcf/d at 932 MW.