The Oil & Gas Value Chain: Oil Sands (2026)

Data as of 19 September 2026. The framework — how bitumen is mined, steamed, blended and upgraded, and what decides whether upgrading pays — is evergreen; the production, cost, capacity and price figures are dated evidence for it, labelled with their reporting year and to be re-sourced when used. Spreads and differentials are annual averages or quarters, never a single day’s print, and no figure here is a valuation of any company. The two producers in Section 5 are hypothetical illustrations. Information only and AI-assisted — see the disclaimer.

Alberta’s oil sands are the third way the world produces oil, and the strangest one: no exploration risk to take, no decline curve to outrun, and a barrel so viscous it cannot enter a pipeline without help. This guide is the oil-sands supplement to the value-chain series — it follows bitumen from the ore face and the steam chamber to the fork every producer reaches, where the barrel is either diluted and sold at the industry’s widest discount or upgraded on site into a light, sweet (low-sulphur) crude. Underneath it sits the question an investor actually has to answer: when does that upgrader pay, and when is it an expensive refinery earning a spread someone else’s pipeline just closed? Every producer behind it can be screened by production, reserves and asset base on Metal Pilot’s upstream oil & gas stock screener.

The headline: the oil sands are not really a drilling business — they are a manufacturing business attached to a reservoir, and that changes almost every number an investor reads.

The Oil & Gas Value Chain series. Part 1: Upstream · Part 2: Midstream · Part 3: Downstream · Part 4: Oilfield Services · Companion: World Oil & Gas Basins · Supplement: Oil Sands

TL;DR & Key Takeaways

• One resource, three production systems. Shallow deposits are mined, deeper ones are melted in place by steam, and the conventional heavy oil next door is pumped cold — Alberta produced close to 3.6 MMbbl/d (million barrels a day) of raw bitumen in 2024.
• The steam is bought, not free. An in-situ project’s steam-to-oil ratio — about 2.65 across Alberta’s SAGD projects in 2021 — is natural gas burned per barrel, which makes a thermal producer a structural gas buyer and its margin partly a bet on the oil-to-gas relationship.
• Every producer reaches the same fork. Bitumen is either blended with 30–35% condensate into dilbit and sold at one of the industry’s widest discounts, or upgraded into synthetic crude. About 48% of Alberta’s raw bitumen was upgraded in 2024, a share the regulator expects to fall toward 44% by 2034.
• The upgrader is half a refinery, and it pays on a spread. Canada’s six upgraders hold 1.41 MMbbl/d of capacity, cost billions to build and earn the light–heavy differential net of gas — and synthetic crude still sold below WTI, the US light benchmark, in six of the eight years to 2025.
• Zero decline is the reward for irreversible capital. Sustaining capital is low and output runs for decades — so the gap between full-cycle and half-cycle breakevens is the industry’s widest.
• The filing names the method. “Oil sands mining and upgrading”, “thermal in situ” and “heavy crude oil” are three cost, capital and price structures in one production table.

Figure 1. The oil sands at a glance

Figure data: Tables 2, 3 and 4, Figure 2 and Section 2.1 of this guide — the sources are printed beneath each.

New to the oil sands? Read straight through from Section 1. Here for the investment case? Jump to Section 5. Every term is defined in the Vocabulary (Section 7).

1. The third production system

Upstream has three ways of getting oil out of the ground, and two of them are drilling. The shale well is fractured and gives up most of its life’s output in its first two years; the offshore well costs a fortune once and then flows for decades. Both are walked end to end in Part 1. The third system does not really drill for oil at all.

Alberta’s oil sands are dug out of the ground or melted in place, and they sit at the far end of nearly every range in upstream: the heaviest barrel, the flattest decline, the widest gap between full-cycle and half-cycle breakevens. They are also the only system that can turn its own product into a light crude on site rather than sell it as it comes out of the ground — an on-lease refinery in all but name, built because the raw product is worth so much less than the market’s reference barrel.

That inversion is what makes the oil sands worth their own guide. A shale producer’s central question is how fast it must drill to stand still; an oil-sands producer has almost no decline to outrun and asks instead what it should sell — a discounted heavy blend, or a light synthetic barrel it has paid billions to manufacture. The geology is settled and the reservoir is not going anywhere. What moves is the spread between those two products, the price of the natural gas that makes the steam, and the cost of the condensate that makes the bitumen flow.

This guide covers the machinery and the choice: mining and steam in Section 2, the upgrader and synthetic crude in Section 3, the conventional heavy oil produced cold next door in Section 4, and a checklist for reading an oil-sands producer in Section 5. The cost stack, the breakevens and the decline arithmetic that these barrels sit inside are in Part 1; Alberta’s place among the world’s producing basins is in World Oil & Gas Basins.

2. Mining & steam

2.1 Mining

Alberta’s oil sands hold bitumen so viscous it is closer to cold asphalt than to crude — 8–11° API, immobile at reservoir temperature — and getting it out is a mining and heat problem rather than a flow problem. Alberta produced close to 3.6 MMbbl/d of raw bitumen in 2024, and the resource behaves nothing like the other two: no decline curve to speak of, no short cycle, and a processing plant attached to the reservoir.

Where the deposit is shallow, it is mined. Shovels and 400-tonne trucks strip the overburden and dig ore that runs 7–13% bitumen by weight; the rest is sand, clay and water. The ore is crushed, mixed with warm water into a slurry and pumped to a primary separation vessel, where bitumen attaches to air bubbles and floats off as a froth that is roughly 60% bitumen, 30% water and 10% fine solids. Froth treatment then adds a solvent to thin it and spin out the rest, leaving a stream that is at least 98% bitumen. The process recovers 85–90% of the bitumen in the ore, recycles 80–90% of its water, and leaves the industry’s most visible liability behind it: tailings — sand, clay, water and unrecovered hydrocarbon — impounded in ponds that have to be reclaimed.

2.2 Steam: SAGD & CSS

Where the deposit is deep, it is melted in place. In steam-assisted gravity drainage (SAGD) a pair of horizontal wells is drilled roughly 5 metres apart, one above the other. Steam injected into the upper well builds a chamber that heats the bitumen until it flows, and gravity drains it into the producer below. The plant’s defining number is the steam-to-oil ratio (SOR) — barrels of steam per barrel of bitumen — because the steam is made by burning natural gas: Alberta’s SAGD projects averaged an SOR of 2.65 in 2021, improved from 2.94 in 2010, and roughly 1.7 gigajoules (GJ) of gas goes into every cubic metre of steam. That makes an in-situ producer a structural buyer of natural gas and a large emitter per barrel, and it makes its margin partly a bet on the oil-to-gas price relationship. The older cyclic steam stimulation (CSS) injects and produces from the same well in cycles, and has gone the other way — an SOR of 5.17 in 2021, up from 3.83 in 2010. In-situ has overtaken mining: in 2021 in-situ produced 1.66 MMbbl/d against mining’s 1.59 MMbbl/d.

How the steam actually gets the oil out. Bitumen does not flow to a well; it has to be made to flow. At the 5–15°C of its own reservoir its viscosity averages around 1 million centipoise — closer to a solid than a liquid — and it only becomes mobile near 50°C, which is why an in-situ operation runs its steam at 200–220°C and needs the rock face itself above roughly 180°C before anything drains. The work is done by condensation: at the edge of the steam chamber, steam meets cold bitumen, condenses, and dumps its latent heat into the rock. The heated bitumen and the water that was steam a moment ago then run down the chamber wall under gravity into the producer below — nothing is sucked or pushed, the well is simply the low point. What arrives at surface is an emulsion of hot bitumen and water, lifted by a downhole pump and piped to a central plant, where the water is separated, treated and sent back to the boilers as steam — up to 90% of it recycled. Getting there takes a start-up phase of its own: for the first weeks or months, steam is circulated through both wells simply to warm the bitumen in the five metres between them, because until that gap softens nothing can drain from one well to the other.

Table 1. The SAGD well pair, drilling to first oil

Source: bitumen viscosity and mobilisation temperatures per Oil Sands Magazine; well-pair geometry per Oil Sands Magazine and the emulsion and water-recycling loop per the Government of Alberta; recovery per Oil Sands Magazine. Durations are indicative and vary widely by project — pad sizes, pair lengths and reservoir quality differ across Athabasca, Cold Lake and Peace River — and describe one well pair rather than a whole project, which ramps pair by pair over years.

2.3 The fork: blend or upgrade

Then comes the fork that decides what the company is selling. Bitumen cannot enter a pipeline as it is, so a producer chooses one of two routes. It can blend it with condensate, an ultra-light liquid produced with natural gas, to pipeline specification — the dilbit of Part 1’s product slate — and sell a heavy, sour (high-sulphur) barrel at one of the widest discounts in the industry while paying for the diluent. Or it can upgrade: run the bitumen through its own coking or hydroconversion plant and sell synthetic crude oil — a light, sweet barrel with no residue, whose process, plumbing and economics are the subject of Section 3. About 48% of Alberta’s raw bitumen was upgraded in 2024, a share the regulator expects to fall to roughly 44% by 2034 as un-upgraded volumes grow faster than upgrading capacity.

Table 2. The oil-sands chain, ore to sold barrel

Source: Alberta’s raw bitumen volumes and the upgraded share per the Alberta Energy Regulator (ST98); ore grade, froth, froth-treatment product and water recycling per Oil Sands Magazine’s mining page, extraction recovery per its extraction page, in-situ recoveries per its in-situ page and the dilbit and synbit blends per its product-streams page; steam-to-oil ratios, the 2021 in-situ and mining volumes and the gas needed per cubic metre of steam per the Canada Energy Regulator; synthetic crude quality per CrudeMonitor. Recovery and SOR figures are project-weighted averages that vary widely by operator and reservoir; the unit chemistry of coking and hydrocracking is covered in the downstream guide.

2.4 What the asset’s shape decides

For an investor the oil sands are the cleanest example of a point this guide keeps returning to: the asset’s shape decides the company’s behaviour. There is no meaningful decline to outrun, so sustaining capital — the spending needed just to hold output flat — is low and production can run for decades — Canadian Natural describes its mining output as zero-decline — but the upfront cost is enormous and the commitment is irreversible, which is why the full-cycle and half-cycle breakevens for oil sands sit further apart than for any other supply source — on the figures Part 1 cites, about $27/bbl WTI-equivalent on average to sustain an existing project in 2025, against $80 or more for a new SAGD project (Part 1, “The cost curve & breakevens”). The upgrader deepens both halves of that trade: it costs billions and it takes the plant down for a turnaround every few years, but it converts the industry’s worst-priced barrel into one priced off WTI — at a premium in some years and a discount in others (Section 3) — and it takes a mine-site producer out of the diluent market. Where a shale company’s key decision is how fast to drill, an oil-sands company’s was made a decade ago and is measured in how reliably the plant runs.

3. Synthetic crude & the upgrader

Section 2 left the oil-sands producer at a fork: dilute the bitumen and sell a heavy, sour blend, or build half a refinery beside the reservoir and sell synthetic crude oil (SCO). The upgrader exists to do one thing: convert the residue — the heaviest, most carbon-rich part of the barrel, which an ordinary refinery cannot use and which makes bitumen so dense — into lighter liquids, strip out the sulphur, and hand the market a crude that flows through any pipeline. What comes out is not a fuel. It is a crude oil, and a strange one: Syncrude Sweet Premium averages 33.0° API and 0.19% sulphur over five years — sweeter than the 0.42% WTI delivery specification (CME Group, NYMEX rulebook), though several degrees heavier than WTI itself — and it has no residue at all, because the upgrader has already converted it.

3.1 Two ways to fix a molecule

Bitumen’s big molecules carry too much carbon for the hydrogen they hold, and there are only two cures: take carbon out, or put hydrogen in. Coking takes carbon out. The residue is heated until its large molecules crack; the lighter pieces boil off as naphtha, distillate and gas oil — the petrol-range, diesel-range and heavier cracking-feed cuts — and the carbon stays behind as solid petroleum coke. It comes in two designs — delayed coking, a batch process in which each drum fills and is emptied over a cycle of about 10 to 16 hours (Suncor, Horizon), and fluid coking, which runs continuously and makes less coke (Syncrude) — and either way it returns 80–90% of the bitumen’s volume. Hydroconversion puts hydrogen in. At Scotford, vacuum residue is mixed with hydrogen and fed into a catalyst-filled LC-Fining reactor, no carbon is rejected, and the volume that comes out is 100% or more of what went in, because the hydrogen makes the product less dense. Whichever route cracks the residue, what it produces is still loaded with sulphur and nitrogen, so every stream then goes through hydrotreaters — the secondary upgrading that makes the crude sweet. Table 3 walks the whole chain.

Table 3. The upgrading chain, bitumen to synthetic crude

Source: the process sequence, delayed and fluid coking, LC-Fining, hydrotreating, hydrogen purity and sulphur recovery per Oil Sands Magazine; coking and hydroconversion volume yields per the Canada Energy Regulator; Syncrude Sweet Premium quality per CrudeMonitor; raw bitumen’s sulphur (about 5%) per Oil Sands Magazine. The chain is generic — every upgrader combines these units differently, and Syncrude runs fluid cokers and an LC-Finer side by side.

3.2 Canada’s six upgraders

It needs a refinery’s worth of plumbing. Canada has six upgraders, with 1.41 MMbbl/d of SCO capacity between them (Table 4), and none of them is a single unit — each is a site built around a supply of natural gas, a hydrogen plant, a sulphur plant and somewhere to put the coke.

Location follows the feed. Three upgraders sit beside the mines that feed them and ship finished SCO south on dedicated pipelines to Edmonton and Hardisty. Scotford sits near Edmonton instead, beside Shell’s refinery, and receives its bitumen as a diluted blend on the Corridor pipeline from the Athabasca mines. Once made, SCO travels as its own product: Enbridge moves it in a separate light synthetic commodity group, and some grades are kept segregated even from other synthetics. And the refinery at the end of the line has to be ready for it. A 2006 UOP study found a typical synthetic crude carried no residue, 50% more vacuum gas oil and 50% more distillate than WTI, and only half the naphtha — and that its cracked, aromatic middle distillates limited how much a refinery could run: as little as 20% of Syncrude’s blend of the time when maximising jet fuel, and about 35% when maximising diesel. SCO’s natural home is the refineries built around it — synthetic crude made up 62% of Alberta’s refinery throughput in 2024 (Alberta Energy Regulator, ST98).

Table 4. Canada’s bitumen upgraders

Source: capacities, locations and the Western Canada total per the Canada Energy Regulator, 2020 data; primary upgrading technology and Scotford’s later capacity per Oil Sands Magazine, 2025; Syncrude’s ownership per the Suncor Energy 2025 annual information form, which puts the base plant at 350,000 bbl/d nameplate; Horizon’s later capacity, the Scotford interest and Corridor feed per Canadian Natural; the Lloydminster feedstock per Cenovus Energy; SCO’s dedicated pipelines per the Suncor annual information form and Pembina Pipeline, and its separate batching per Enbridge; the refinery-fit figures per UOP, 2006. Capacities are the regulator’s 2020 figures; Canadian Natural has since put Horizon at about 264,000 bbl/d, and Oil Sands Magazine puts Scotford at about 328,000 bbl/d. n/d marks a technology not confirmed from a published source. Nexen’s Long Lake upgrader was shut after a 2016 fire, and the Sturgeon refinery processes bitumen into refined products rather than SCO, so neither is listed.

3.3 What upgrading costs

Upgrading is capital-hungry, and the Canada Energy Regulator blames “skyrocketing capital costs” for the fact that no one builds it any more (Section 3.4). Table 5 collects what is disclosed, from Horizon’s integrated mine and upgrader at C$88,200 per flowing barrel — capital per barrel a day of SCO capacity — in 2008 to the Sturgeon bitumen refinery, a different plant but the closest recent build, at about C$127,000 per barrel a day of blended feed, or C$200,000 per barrel a day of bitumen. The running costs are harder to isolate, because no producer reports its upgrader apart from its mine; the whole mined-and-upgraded barrel sits in Part 1’s operating-cost table.

Table 5. What upgrading costs

Source: Horizon Phase 1 per Canadian Natural, 6 November 2008 (Form 6-K, Exhibit 99.3); Sturgeon’s final cost and capacity per the Alberta Petroleum Marketing Commission, annual report to March 2021, its bitumen share of the blend per Fluor, the builder, and the original estimate per North West Redwater, December 2013; the Voyageur impairment per Suncor Energy, Q4 2012, its cancellation per Suncor Energy, March 2013, and Total’s loss per TotalEnergies; natural gas cost per Canadian Natural’s 2025 management’s discussion and analysis (MD&A) and its operating cost per its 2025 fourth-quarter report; Syncrude cash operating cost per Suncor Energy, Annual Report 2025. The per-barrel capital of Sturgeon and its overrun are this report’s arithmetic on the published figures; capital figures are nominal and not adjusted for inflation.

Two costs never reach Table 5: the coke and the carbon. The coke is the least visible problem and one of the largest: Alberta’s upgraders made 8.6–10.1 million tonnes of it a year over 2013–2017, burned about 10% as fuel in 2016 and stockpiled the rest, so the inventory grew from 84 to 113 million tonnes in four years (Environment and Climate Change Canada, draft screening assessment). Suncor retired its coke-fired boilers at the base plant in late 2024, replacing them with an 800 MW cogeneration plant, which makes power and steam together from natural gas (Suncor Energy, 2025 annual information form). The hydrogen plants are the carbon problem: Shell’s Quest project captures up to 1.08 million tonnes of CO₂ a year from Scotford’s hydrogen units — about 35% of the upgrader’s CO₂ — and has stored more than 10 million tonnes since 2015 (Government of Alberta).

3.4 Does it always pay?

It pays on a spread. An upgrader earns the gap between what synthetic crude sells for and what the same bitumen would have fetched as a blend, and it has to pay its gas, its operating cost, the barrels it loses to coke and the return on its capital out of that gap. The regulator describes the two sides plainly: SCO is tied to WTI, while bitumen’s value is linked to WCS (Western Canadian Select, the heavy benchmark) and the price of diluent — so the upgrader’s margin is, in effect, the light–heavy differential. That spread moves a great deal. On Suncor’s benchmark prices, synthetic crude at Edmonton was worth US$20.60/bbl more than WCS in 2023, US$14.10 in 2024 and US$10.80 in 2025 — a narrowing that was mostly the heavy discount itself: WCS’s discount to WTI fell from US$18.65 in 2023 to US$14.73 in 2024 (Alberta Energy Regulator, ST98) and US$11.10 in 2025 (Canadian Natural), after the Trans Mountain pipeline expansion to the Pacific coast opened in May 2024. Inside Canadian Natural, the realised SCO price of C$86.41/bbl in 2025 stood C$20.18 above the value it placed on its bitumen for royalty purposes, down from C$25.35 a year earlier — before the upgrader’s own costs, and not on a like-for-like volume basis.

Nor does synthetic crude reliably sell above WTI. Figure 2 shows it at a discount in six of the eight years to 2025, and three things hold it there, all belonging to the barrel or its address rather than to the market’s mood. It is priced at Edmonton, not at Cushing, Oklahoma, where WTI is delivered, so it carries the cost of reaching US refiners. It is several degrees heavier than WTI, with half WTI’s naphtha and a cracked, aromatic middle of the barrel that caps how much of it a refinery can run (Section 3.2) — which is why Alberta’s own refineries are its natural buyers. And when those refineries and the pipelines out of the province are full, the marginal barrel has to be discounted to find a home — which is where the largest swings come from: in the fourth quarter of 2018, when Alberta’s pipelines were full, SCO sold US$21.35/bbl under WTI; in the second quarter of 2026 it sold US$8.37/bbl over it. An upgrader protects its owner against a wide heavy-oil discount, but not against a glut of light oil in Alberta or a shortage of pipe out of it.

Figure 2. Synthetic crude’s premium or discount to WTI, 2018–2025

Chart source: Suncor Energy’s “SYN-WTI” differential from its annual MD&A (management’s discussion and analysis) benchmark tables — 2019 for 2018–2019, 2022 for 2020–2022 and 2025 for 2023–2025, chosen as one publisher’s consistent series. The SCO–WCS spreads in the text are this report’s arithmetic on the same 2025 table (WTI plus the SYN–WTI differential, less WCS at Hardisty); the 2018 and 2026 quarterly figures per Canadian Natural, Q4 2018, and Canadian Natural, Q2 2026; the bitumen royalty value per Canadian Natural’s 2025 MD&A; the WTI and WCS linkage per the Canada Energy Regulator. Bars share one scale on both sides of zero.

That is why nobody is building upgraders. The Canada Energy Regulator puts it down to skyrocketing capital costs and shrinking light–heavy differentials, and the Alberta Energy Regulator’s ten-year outlook assumes no new upgrading facilities at all, so the upgraded share of Alberta’s bitumen drifts down as raw production grows (Section 2.3). The record of the last two decades makes the same point: Suncor abandoned the Voyageur upgrader in 2013, saying market conditions had “changed significantly” since 2010; Long Lake’s upgrader was shut after a 2016 fire; and Sturgeon’s overrun is in Table 5. Alberta’s attempt at a cheaper middle way — a C$1 billion programme of grants and loan guarantees announced in 2018 for partial upgrading, which processes bitumen just enough to cut the diluent it needs — was cancelled in October 2019 (University of Calgary School of Public Policy), and none of the technologies has reached commercial readiness; the most advanced are at demonstration stage (Energies, 2023).

So upgrading pays under three conditions, and investors should check all three. The plant must already exist: the capital is sunk, and an existing upgrader is judged on its half-cycle margin, not on what it cost. The light–heavy spread must be wide enough, net of gas: a narrow WCS discount, as after 2024, thins the margin however well the plant runs. And it must run reliably, because an upgrader’s economics live in its utilisation — a turnaround or an outage removes the premium barrels while the fixed costs stay. When all three hold, an integrated miner-upgrader sells a light sweet barrel from the world’s heaviest resource and stays out of the diluent market; when they do not, it owns an expensive, carbon-intensive refinery earning a spread that someone else’s pipeline just closed.

4. Heavy oil & cold production

“Heavy oil” and “oil sands” are used interchangeably in conversation and mean different things in a filing. The dividing line is not the label but the viscosity, because viscosity decides whether a pump alone can do the job. Alberta’s bitumen at a million centipoise cannot be pumped at any rate; the conventional heavy oil of the Lloydminster region — broadly 10–20° API and above 100 cP — is thick, but it moves. Most of it is produced cold, with no steam anywhere on the lease.

The method that made that rock economic is cold heavy oil production with sand (CHOPS), and its trick is counter-intuitive: the operator deliberately produces the formation’s sand along with the oil rather than screening it out. Sand leaving the formation opens high-permeability channels — “wormholes” — that radiate from the well and let far more oil reach it, and the foamy oil that comes with dissolved gas helps drive the flow. The rate difference is the whole story: a heavy-oil well that keeps its sand out produces perhaps 0.5–5 m³ a day, while one producing sand can reach 15–50 m³ a day. The equipment is chosen to match — a progressive cavity pump that can turn through a sand-laden fluid rather than a conventional rod pump — and the cost is that the sand has to be separated, trucked and disposed of for the life of the well.

What cold production does not do is recover much. A CHOPS well drains a small share of the oil around it, which is why the region is the natural home of the follow-on methods in Part 1, “Recovery factors & enhanced oil recovery”: polymer flooding, where the injected water is thickened so it pushes heavy oil instead of slipping past it, and small-scale thermal projects on what CHOPS left behind. And heavy oil is steamed elsewhere in the world without being anything like an oil sand — California’s heavy-oil fields are the centre of US thermal recovery, running steam into conventional reservoirs that have been producing for a century.

Table 6. What the heavy end actually is, and how it is produced

Source: the CHOPS definition, its 10–20° API / 100 cP range and the sand-versus-no-sand well rates per the Government of Alberta CHOPS report; bitumen viscosity per Oil Sands Magazine; the US thermal-recovery concentration in California per the U.S. Department of Energy. API and viscosity bands are conventions that overlap at the edges, and several of these methods are used on the same field at different stages of its life.

For a reader of company reports the practical point is that the production line item names the method, and the words are standardised. “Oil sands mining and upgrading” is a mine and a plant; “thermal in situ” is steam; “heavy crude oil” on a Canadian filing is almost always pumped cold; “steamflood” or “cyclic steam” outside Alberta is thermal recovery on a conventional field. Canadian Natural, for one, reports mined synthetic crude, thermal bitumen and primary heavy crude as separate lines in the same production table, which is exactly why the mix matters more than the barrel-of-oil-equivalent (boe) total (Part 1, “Units & reading a production number”). The economics differ as sharply as the machinery: a cold heavy-oil well is cheap, quick and recovers little, while a thermal project is a decade-long capital commitment that recovers most of the reservoir — and both sell into the same discounted heavy market (Sections 2.3 and 3.4).

5. How to read an oil-sands producer

The machinery above resolves into a small set of questions, and a Canadian producer’s own disclosure answers most of them. The production table names the method, the sales mix names the product, and the operating statistics name the gas exposure — so an oil-sands company can be read without a single forecast.

Start with what the company physically operates, because that decides everything after it. A mine and an upgrader are a manufacturing complex with decades of flat output and an enormous sunk cost; a SAGD project is a steam plant whose margin moves with the gas price; a cold heavy-oil well is Section 4’s cheap, low-recovery option. Most of the large producers run more than one, and the mix matters more than the barrel count.

Then ask what the barrel is priced against. An upgraded barrel is tied to WTI; a blended barrel is tied to WCS less the cost of the condensate inside it. That single difference explains why two producers with identical volumes can report netbacks — the price realised per barrel less royalties, transport, diluent and operating costs — a world apart, and why a narrowing light–heavy spread hurts the upgrader’s owner while helping the blender.

Table 7. The oil-sands checklist

Table note: a reading framework, not investment advice — each row points to the section of this guide that explains it.

The checklist at work — two hypothetical producers. Producer A is an integrated miner-upgrader that sells synthetic crude; Producer B is a SAGD producer that sells dilbit. Both are hypothetical illustrations built from this guide’s figures, not real companies, and the gas prices are round numbers chosen to show the sensitivity, not a forecast.

Table 8. The checklist applied to two hypothetical producers

Source: hypothetical illustration built from Tables 2 and 5, Section 2.2 (SOR and 1.7 GJ of gas per m³ of steam) and Section 3.4 (the Suncor benchmark spreads, whose heavy-discount share of the narrowing is this report’s arithmetic on the same table); the 2025 benchmark gas price of C$1.76/GJ per Canadian Natural’s 2025 MD&A; 1 barrel = 0.159 m³. Gas prices are round assumptions, not quotes. Not the profile of any company.

The two can report the same volumes and still move in opposite directions. A narrowing spread takes margin from A and helps B; a rise in the gas price lands on B’s steam bill, and twice as hard if its reservoir needs CSS-level steam. That is what the checklist’s first rows are for: they tell you which way each company’s netback will move before the quarter is reported.

Read in that order, the checklist does one thing: it separates the questions the asset has already answered from the ones the market answers every day. Geology, method and plant are fixed; the spread, the gas price and the diluent cost are not. A producer whose barrels are already upgraded and whose plant is running well is levered to the light–heavy differential; one selling dilbit is levered to pipeline space out of Alberta and the price of condensate. To take that reading to the company level — screening every producer by production, reserves, netback and asset base — use Metal Pilot’s upstream oil & gas stock screener.

6. Summary

The oil sands are the industry’s third production system, and the one that behaves least like the other two. Bitumen at 8–11° API does not flow, so it is either dug out of shallow deposits and washed free with warm water, or melted underground by steam and drained to a producing well five metres below the injector. The cost of that heat is the defining number of an in-situ project: the steam-to-oil ratio is natural gas burned per barrel, which makes a thermal producer a structural gas buyer and a large emitter, and hands it a margin that moves with the oil-to-gas relationship as much as with crude itself.

Whatever the method, every producer reaches the same fork. It can blend the bitumen with condensate to pipeline specification and sell a heavy, sour blend at one of the industry’s widest discounts, paying for the diluent and shipping it both ways — or it can upgrade, running the bitumen through its own coker or hydroconversion plant and selling a light, sweet synthetic crude with no residue at all. Canada has six such plants and 1.41 MMbbl/d of capacity, and nobody is building more: capital costs have risen, the light–heavy spread has narrowed, and the regulator’s ten-year outlook assumes no new upgrading facilities at all.

That is the argument an investor has to settle. An upgrader pays when the plant already exists, the spread is wide enough net of gas, and the plant runs reliably — and synthetic crude still sold below WTI in six of the eight years to 2025, so the premium is neither automatic nor permanent. Beside the oil sands proper sits conventional heavy oil, produced cold with the sand deliberately let in, cheap per well and thin on recovery, selling into the same discounted market. The practical skill is reading which of these a company actually owns, because the production table names the method and the method names the economics.

To move from this map of the resource to the companies producing it — every one screened by production, reserves and asset base — explore Metal Pilot’s upstream oil & gas stock screener.

Back to the series → Part 1: Upstream for the wellhead mechanics, cost stack and breakevens these barrels sit inside — or see where Alberta ranks among the world’s producing basins in World Oil & Gas Basins.

7. Vocabulary

Table 9. Oil sands and heavy oil terminology

Source: definitions compiled from the process and market descriptions in Sections 2–4 and their cited sources — above all Oil Sands Magazine, the Canada Energy Regulator and the Alberta Energy Regulator.

8. Sources, methodology & disclaimer

8.1 Sources, methodology & data vintage

Regulators and government. Alberta’s raw bitumen volumes, the upgraded share and the ten-year outlook per the Alberta Energy Regulator (ST98); in-situ trends and steam-to-oil ratios per the Canada Energy Regulator, and upgrader capacities, yields and the WTI–WCS linkage per its tour of Canada’s oil sands upgraders; the CHOPS definition and well rates per the Government of Alberta; petroleum-coke production and inventories per Environment and Climate Change Canada; the Sturgeon refinery’s final cost per the Alberta Petroleum Marketing Commission.

Industry and market data. Process descriptions and recoveries — mining, extraction, froth treatment, in-situ methods and upgrader design — per Oil Sands Magazine’s technical pages (mining, in-situ, upgrading); crude qualities per CrudeMonitor five-year averages.

Company disclosures. Capacities, costs, capital and realised prices from Suncor Energy, Canadian Natural, Cenovus Energy and Shell Canada filings, results releases and annual information forms, 2008–2026, each cited at the table or figure that uses it. Company names identify operators and are not endorsements; no position in any named company is implied.

Methodology & data vintage. Process mechanics, the mining and in-situ routes, the upgrading chain and the fork between blending and upgrading are treated as evergreen; every volume, capacity, cost and spread carries its reporting year and should be re-sourced before use. Where the regulator states the upgraded share of Alberta’s bitumen on two bases — 48% and 41% for 2024 — this guide quotes the executive summary’s figure and names it as such. Upgrader capacities are the regulator’s 2020 data and are stated as nameplate, not throughput. Project durations in Table 1 are indicative and vary widely by reservoir and pad design. The synthetic-crude differentials in Figure 2 are one publisher’s consistent annual series, and the SCO–WCS spreads quoted in the text are this report’s arithmetic on that same benchmark table rather than a published series. The WCS–WTI differentials are the regulator’s for 2023–2024 and Canadian Natural’s for 2025; Canadian Natural reports the same US$14.73 for 2024. No figure here is a live quote or a valuation.

Data as of 19 September 2026. Intended update cadence: annually, after the Alberta Energy Regulator’s ST98 outlook and the year-end company reports.

8.2 Disclaimer & disclosure

This guide is for informational purposes only and is not investment advice, a recommendation, or an offer to buy or sell any security. Oil sands production, capacity, cost and price figures are estimates as of the dates stated and change continually with prices, regulation, maintenance and corporate transactions; readers should verify current figures against primary sources before acting, and consider consulting a licensed financial adviser. Company names are given to identify operators, not to endorse them, and no position in any named company is implied. This report was prepared with AI assistance: figures were sourced from the regulators and companies cited above and reviewed for accuracy, but readers should independently verify any number they rely on. Past production, cost and spread patterns are not a guarantee of future results.