The Inquiry Vol. I·Vol. II·Vol. III·Vol. IV·Vol. V·The Survey September 2026  ·  EN
Investigative Economics
Four Conversion Platforms, Three Sales, and the One Nobody Would Buy

Built
to Be
Sold.

Canada produces 6.1 million barrels of oil a day and can refine 1.9. It has built the plants that would close that gap — a smelter fed by its own river, a world-scale fertilizer works, a refinery, a pipeline — four times in seventy-five years. It sold three of them, at prices it announced, and kept the fourth only because no buyer could be found. This is the record of those four, what each cost, and what each fetched.

Subject1951 – 2026
MechanismThe sale as the return
Preceding volumeConsequence Without Conviction
Read time12 minutes
Volume V A record and an instrument. The record is four energy-conversion platforms built in Canada since 1951, dated, and what happened to each. The instrument prices converting at home in the only currency Canada has actually paid for it — the one refinery it finished.
i. The Water

Canada converts
one barrel in four
and ships the rest.

Four numbers open this, and they are all Canadian.

Canadian oil production, 2025
6.1Mb/d
Oil sands 3.5, conventional 1.1, NGLs 0.7, condensate 0.6, offshore 0.2.
Canadian refining capacity
1.9Mb/d
Fourteen refineries, two of them asphalt plants. The whole country.
What those refineries actually ran
1.6Mb/d
Ninety per cent of capacity in 2025. There is no idle conversion to switch on.
Crude exported, unconverted
4.0Mb/d
About ninety per cent of it to the United States. Sixty per cent of everything America imports.

Canada converts roughly one barrel in four of what it pulls out of the ground and ships the rest as crude, overwhelmingly to one customer. The refining that would close that gap is not idle capacity waiting to be switched on; there is none spare. Closing it means building.

Canada has built something to close that gap four times since 1951: a smelter fed by a river it re-routed, a world-scale fertilizer works, a refinery, and a pipeline. What follows is the record of those four.

Canada has built the plants that would close the gap. It sold three of them, at prices it announced, and kept the fourth only because no buyer could be found.
ii. The Record

Four platforms,
dated, and what
happened to each.

Each entry below is public and dated. Four platforms, from the first water licence to the divestment now under way.

Leg one
The platforms get built.
1951
British Columbia signs over land and water rights on the Nechako to Alcan. The province contributes the resource; Alcan contributes the capital. What follows is the largest privately funded construction project in Canadian history: a dam that reverses a river, a tunnel through a mountain, a powerhouse inside it, and a town.Resource for capital
3 Aug 1954
First aluminum is poured at Kitimat, powered by Kemano. Hydroelectricity in, metal out, at world scale, five years after the first survey.The architecture works
1992
SaskFerco is commissioned at Belle Plaine: Cargill 50 per cent, Investment Saskatchewan 49 per cent, Citibank Canada 1 per cent. Saskatchewan gas goes in; ammonia, urea and UAN come out. The province is a near-equal owner of a world-scale nitrogen plant.Province at 49%
Feb 2011
Alberta signs the Sturgeon Refinery agreement. The Alberta Petroleum Marketing Commission supplies 75 per cent of the bitumen, carries 75 per cent of the tolls and 75 per cent of the subordinated debt — and holds 25 per cent of the votes. An internal evaluation team had flagged the disproportionate risk in April 2010.75% risk, 25% say
Leg two
The platforms get sold.
23 Oct 2007
Alcan joins Rio Tinto for US$38.1 billion, US$38.7 billion with fees — the largest all-cash mining deal to that date, and later widely called the worst. Canada keeps the river, the dam, the tunnel, the smelter and the workers. It stops keeping the decision about any of them.Sold
1 Oct 2008
The sale of SaskFerco to Norway’s Yara International closes in a $1.6-billion transaction. Investment Saskatchewan’s 49 per cent returns $783 million to the province, which announces it. Nobody took it. Saskatchewan sold it and booked the gain.Sold for $783m
Jun 2014
Six months after Sturgeon’s estimate moves from $5.7 billion to $8.5 billion, Alberta discloses that its processing-fee commitment has gone from $19 billion to $26 billion. The plant is not built yet.Before first oil
Leg three
The one that could not be sold.
2018
Ottawa buys the Trans Mountain pipeline for $4.5 billion to rescue an expansion estimated at $7.4 billion in 2017.Public purchase
Feb 2018
Alberta’s Auditor General reports on the Sturgeon agreement: the province carries the risk “as if it were building the refinery,” on 75 per cent of the tolls. High benefit, high risk. A second report follows in November 2020.Auditor General
2020
The marketing commission’s own annual report values the Sturgeon position at a negative $2.52 billion net present value. Sturgeon reaches full operation in May of the same year.Negative NPV
6 Jul 2021
Alberta takes 50 per cent of the equity in Sturgeon. It could not exit, so it bought in. Canadian Natural Resources holds the other half.No exit available
May 2024
Trans Mountain goes into service at a final cost of $34.2 billion, against $7.4 billion estimated seven years earlier. Its function is to move unconverted crude to tidewater.$34.2bn
Mar 2024
Sturgeon’s total future toll commitment stands at $37.3 billion, Alberta’s share about $28 billion. The commission paid $930.7 million in tolls that fiscal year. Most of the debt repayment does not start until 2032.$28bn public

The bottleneck was never Canada’s inability to build. Canada built Kitimat in five years with a river and a mountain in the way, and the largest privately funded construction project in its history came out the other side. It built Belle Plaine and owned forty-nine per cent of it. It built Sturgeon. It built a thirty-four billion dollar pipeline.

The bottleneck is what Canada does once the thing is standing.

iii. The Sale

Three sales, one
failed sale, and what
each one fetched.

In three of the four cases the platform changed hands by sale — negotiated in the open, at a disclosed price, announced by the seller. In the fourth it did not change hands because no buyer could be found.

Alcan → Rio Tinto, 2007
$38.1bn
US dollars, all cash. A shareholder vote, not an expropriation.
Saskatchewan’s 49% of Belle Plaine, 2008
$783m
Announced by the province as a payoff on the investment.
Trans Mountain, built 2024
$34.2bn
Bought for $4.5 bn in 2018. Expansion estimated at $7.4 bn in 2017.
What analysts say it is worth
15–25bn
Divestment under way in two phases, Indigenous equity first.

Put the four beside each other and the mechanism is plain. The sale is the return. Canada does not build conversion platforms as capacity to be held; it builds them as assets to be exited. When the exit clears a gain, the province issues a news release. When it clears a nine-to-nineteen-billion-dollar loss, the exit happens anyway, because holding was never the plan.

Sturgeon is the exception that proves it. Alberta ended up owning half a refinery in July 2021 not because it decided to retain strategic conversion capacity, but because there was no buyer and no way out. The single case where Canada holds a conversion platform outright is the case where Canada could not sell.

Qatar Fertiliser was founded in 1969 on the same structure — a joint venture with foreign shareholders, turning cheap gas into a transportable product. It now runs six plants, 3.8 million tonnes of ammonia and 5.6 million tonnes of urea a year, and is the largest single-site urea exporter in the world. Industries Qatar owns one hundred per cent of it.

A fifth is being attempted now, at Belle Plaine, on the same site with the same gas and the same product. Genesis Fertilizers is a $2.89-billion nitrogen complex, explicitly farmer-owned to answer the ownership question directly. Farmers have put in $33 million. That is one point one per cent of the project. Twenty-four million of it is already spent, twenty on engineering and four on land, and the company is seeking government support for the rest.

Its letter of intent for natural gas supply, product sales and carbon credits is with Gunvor, one of the largest commodity trading houses on earth. Final investment decision was targeted for early 2026, construction for the second half of 2026, production for 2029.

It is the current attempt at Canadian ownership of a conversion platform. At the stage where ownership is settled it stands at 1.1 per cent farmer equity, with gas procurement and product marketing contracted to a Geneva trading house.

iv. The Price

The Sturgeon Rule:
what converting
at home costs.

There is a reason Canada keeps arriving at the same decision, and it is a number. Converting at home has a price, and Canada is the only country that has recently paid it in public and written that price down. The Sturgeon Refinery at Redwater takes 80,000 barrels a day of bitumen and makes diesel. It was estimated at $5.7 billion, capped at $6.5 billion, revised to $9.4 billion in May 2017, and came in near $10 billion. As of March 2024 the total future toll commitment across its life was $37.3 billion, of which Alberta’s marketing commission carries about $28 billion, having paid $930.7 million in 2023–24 alone.

That gives two hard conversion rates: $125,000 of capital per daily barrel, and $466,250 of thirty-year tolls per daily barrel. Move the slider and converting at home gets priced in the only currency Canada has actually paid.

The Sturgeon Rule · calibrated on Redwater, 2011–2024

Convert 207,000 barrels a day of Canadian crude at home instead of exporting it raw.

80,000One Sturgeon
207,000Already sent to China
1,900,000Every Canadian refinery
4,000,000All crude exports
Sturgeon refineries
required
2.6×
At 80,000 barrels a day each.
Capital, at Redwater’s
realised cost
$25.9bn
Thirty-year toll
commitment
$96.5bn
Capital rate: $10 bn ÷ 80,000 b/d = $125,000 per daily barrel. Toll rate: $37.3 bn ÷ 80,000 b/d = $466,250 per daily barrel over thirty years, of which Alberta’s share is 75%. Both figures are Redwater’s, as built and as filed.

Read it at the low end first, because that is where it works. One Sturgeon is two per cent of the crude Canada ships out. Belle Plaine is one plant. Alouette is one smelter. At that scale this is not a theory, it is a construction schedule, and every working Canadian example is sitting at exactly that scale today.

Read it at the top and it stops being a programme. Converting the 4 million barrels a day Canada currently exports raw is fifty Sturgeons, half a trillion dollars of capital, and a toll obligation of $1.87 trillion across thirty years. That capital figure is thirteen times what Rio Tinto paid for all of Alcan. It is fifteen Trans Mountains.

Canada priced conversion at Redwater, and the Auditor General wrote it up twice. That is the number every subsequent decision has been made against.

Which puts the real question at plant scale rather than national scale. One plant is affordable and always has been. The record is about what happens to the ones that get built.

Canada built Kitimat in five years and sold Alcan in 2007.
It owned forty-nine per cent of Belle Plaine and took $783 million for it in 2008.
It built Sturgeon, could not sell it, and carries $28 billion in tolls.
It spent $34.2 billion on a pipeline worth $15 to $25 billion, and is selling that too.
Qatar was handed the same deal in 1969 and still owns all of it.
The question was never who captures the surplus. It is who agreed to stop holding the thing that makes it.

Sources

  1. Canadian production and refining: Canada Energy Regulator, “Canada sets new record in crude oil production in 2025” — 6.1 million b/d, oil sands 3.5, conventional 1.1, NGLs 0.7, condensate 0.6, offshore 0.2; and “Canadian crude refinery runs held steady in 2025” — fourteen refineries, 1.9 million b/d of capacity, 1.6 million b/d run at ninety per cent utilisation, about 400,000 b/d of refined products exported.
  2. Crude export destination: CER, Overview of Canada–U.S. Energy Trade, and The Hub, 5 June 2026. About 4 million b/d to the United States, roughly sixty per cent of American crude imports.
  3. The Sturgeon Refinery: project history and cost sequence — $5.7 bn estimate, $6.5 bn cap in 2011, $9.4 bn in May 2017, near $10 bn final, 80,000 b/d, agreement signed February 2011 after an April 2010 internal warning on disproportionate risk, Alberta taking 50 per cent of the equity on 6 July 2021. Toll escalation from $19 bn to $26 bn reported June 2014. Auditor General of Alberta reports February 2018 and November 2020; APMC 2020 annual report carrying a negative $2.52 bn net present value. March 2024 total future toll commitment of $37.3 bn, Alberta’s share about $28 bn, with $930.7 m paid in 2023–24 and debt repayment largely deferred to 2032. Contemporaneous coverage: CBC News, Troy Media.
  4. Trans Mountain: purchased for $4.5 bn in 2018 against a $7.4 bn expansion estimate from 2017; in service May 2024 at $34.2 billion; analyst valuations of $15–25 bn and the two-phase divestment beginning with Indigenous equity, per CBC News and Benefits and Pensions Monitor. Federal purchase terms at CER.
  5. Alcan: BC land and water rights agreement 1951, Kemano energised 15 July 1954, first pour at Kitimat 3 August 1954, the largest privately funded construction project in Canadian history — Rio Tinto, 70 years at Kitimat, Kemano Generating Station. Acquisition completed 23 October 2007 at US$38.1 billion, US$38.7 bn with fees.
  6. SaskFerco and Belle Plaine: commissioned 1992 with Cargill 50 per cent, Investment Saskatchewan 49 per cent, Citibank Canada 1 per cent; Cargill’s interest passing to Mosaic in 2004; sale to Yara International closing 1 October 2008 in a $1.6 bn transaction returning $783 million to the province. See also The Encyclopedia of Saskatchewan and The Western Producer.
  7. QAFCO: six plants at Mesaieed, 3.8 Mt of ammonia and 5.6 Mt of urea a year, the world’s largest single-site urea exporter at up to fourteen per cent of world supply, founded 1969 as a joint venture and now wholly owned by Industries Qatar — QAFCO, plants, Gulf Times.
  8. Genesis Fertilizers: the proposed complex; $2.89 bn project cost, $33 m raised from farmers of which $24 m spent — $20 m on front-end engineering and $4 m on land — and government support sought, per The Western Producer. Letter of intent with Gunvor covering natural gas supply, product sales and carbon credits per Top Crop Manager. Final investment decision targeted early 2026, construction second half of 2026, production 2029.