The Oil & Gas Value Chain, Part 1: Upstream (2026)

Data as of 26 August 2026. The framework — how upstream earns, what a barrel costs, how reserves are graded — is evergreen; the production, reserve, cost and deal figures are dated evidence for it, labelled with their reporting year and to be re-sourced when used. Prices and breakevens are periods or averages, never a single day’s print, and the one worked barrel (Table 9) is a hypothetical illustration, not a valuation of any lease or company. Information only and AI-assisted — see the disclaimer.

Every barrel of crude and every cubic foot of gas the world burns begins in the same place: a hole in the ground, drilled on a geologist’s bet, into a rock that has held hydrocarbons for tens of millions of years. Upstream is the first third of the oil and gas value chain — exploration and production, universally shortened to E&P — and it is the only part that actually creates the molecule the other two thirds move and refine. That makes it behave nothing like the pipelines and refineries most energy investors also own: upstream does not sell a service or a processing margin, it sells the commodity itself, at whatever the world will pay that day. That single fact is the whole investment case and the reason upstream is the highest-beta, most violently cyclical link in the chain — and every producer behind it can be screened by production, reserves and asset base on Metal Pilot’s upstream oil & gas stock screener.

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

• What it is: the finding-and-producing end of oil and gas — exploration, appraisal, development and production of crude oil and natural gas. Upstream owns the resource in the ground and sells the raw molecule at the wellhead.
• How it earns: as a price-taker — revenue is volume times the commodity price, minus the cost of lifting and finding the barrel, with no fee and no spread, which makes upstream the most price-sensitive part of the chain.
• The scale: the United States pumped a record 13.6 MMbbl/d of crude and 118.5 Bcf/d of marketed gas in 2025, out of roughly 106 MMbbl/d of world liquids output.
• The scarcity signal: shortage shows up as a rising price and a rising breakeven, not a spread or a basis blowout — and because a new Permian well starts near 1,100 boe/d (derived from EIA data) and, on a modelled decline curve, gives up roughly a quarter of its lifetime barrels inside twelve months, the industry must re-drill continuously just to stand still.
• The equity behaviour: upstream amplifies the commodity — the mirror image of midstream, levering the barrel in both directions: spectacular in a boom, brutal in a bust.
• The biggest swing factor: capital discipline, consolidation and the peak-demand debate — a roughly $300 bn US merger wave in 2023–24 (ExxonMobil–Pioneer at $64.5 bn) has remade the Permian, the IEA and OPEC disagree on whether oil demand plateaus around 2030 or keeps climbing to 2050, and US gas is being pulled toward liquefied natural gas (LNG) exports and data-centre power.

Numbers to remember (upstream at a glance)

Figure 1. Upstream oil & gas at a glance

Figure data: U.S. EIA, with world liquids output from the EIA’s Short-Term Energy Outlook, August 2026, reserves from the Oil & Gas Journal end-2024 survey, the NOC share from NRGI, 2019, Dallas Fed, Wood Mackenzie and deal disclosures; see Sections 4.2, 5.2, 6.1 and 7.4.

How to read this guide: new to the sector? Read straight through from Section 1. Here for the investment case? Jump to Section 7. Every term is defined in the Vocabulary (Section 10).

1. Upstream & the market basics

1.1 The E&P lifecycle

Upstream is the first third of the hydrocarbon value chain. It finds hydrocarbons and produces them; midstream moves, processes and stores what upstream lifts; downstream refines it into fuel and sells it. Where midstream earns a fee and downstream a margin, upstream sells the raw commodity at the wellhead and takes whatever the market pays — which is why it behaves like a leveraged bet on the oil and gas price rather than like a toll road or a factory.

It is best understood as a lifecycle, because a barrel of reserves passes through a fixed sequence of stages before it becomes cash, each with its own risk and capital profile:

• Exploration — the geological bet: seismic surveys and a handful of expensive wildcat wells test whether a prospect holds hydrocarbons. Most find nothing, and a small company can be made or destroyed on a single result.
• Appraisal — wells map how big a discovery is and whether it can be produced economically; this is where a resource starts to become a reserve.
• Development — the heavy-capital stage of production wells, platforms and processing plant; a large offshore project can take a decade and tens of billions of dollars between the final investment decision (FID) and first oil.
• Production — the payback stage, in which the field flows and the operator spends to keep it flowing against natural decline.
• Decommissioning — plugging wells and removing infrastructure at the end of field life, an obligation carried on the balance sheet as an asset-retirement obligation (ARO) from the day production starts.

Figure 2. The upstream lifecycle, prospect to plug

Source: standard oil & gas value-chain structure (upstream–midstream–downstream) and E&P project lifecycle per U.S. EIA primers; conceptual diagram.

A company’s position on that lifecycle tells an investor most of its risk: an explorer is a binary bet on drilling results, a developer a bet on execution before first cash, and a producer a bet on the commodity price and on how cheaply it can replace what it pumps. This guide runs from the prospect to the wellhead, where the barrel is handed to midstream.

1.2 The product slate & crude quality

“Oil and gas” is shorthand for a continuum that runs from gases lighter than air to a tar that sinks in water, with a few non-hydrocarbon streams riding along, and density puts all of it in order — among the hydrocarbons it is also a fair first read on value. Gas is sold by the Mcf (thousand cubic feet) at a fraction of oil’s price for the same energy, the natural gas liquids (NGLs) stripped out of it — ethane, propane, butanes — sit between gas and crude, the middle liquids turn into fuels with the least work, and every step down in gravity past that point costs a discount, diluent and processing. Table 1 sets the slate out on one basis, density at 15 °C and atmospheric pressure; how each product is priced is the subject of the crude oil and natural gas market guides.

Two columns carry the argument. API gravity, the industry’s density scale for liquids (explained below), puts its 10° mark at the density of water, so raw bitumen sits on or just past the line and must be blended before it can move. And the boe column counts every liquid hydrocarbon barrel for barrel and gas at 6 Mcf to the barrel (Section 1.4), while helium, carbon dioxide, water and sulphur do not count at all, however much money they make or cost.

Table 1. The upstream product slate, lightest to heaviest

Source: gas and NGL densities at 15 °C and 101.325 kPa per the NIST Chemistry WebBook — the bracketed liquid densities are the saturated liquid at the same temperature, and the butanes row is normal butane; natural gasoline at 0.1055 t/bbl per U.S. EPA; five-year average density and gravity of the Edmonton condensate blend (CRW), Synthetic Sweet Blend and Western Canadian Select per CrudeMonitor, CrudeMonitor and CrudeMonitor; bitumen’s 8–11° gravity and its sitting heavier than water per Oil Sands Magazine; upgrader yields and SCO’s WTI pricing per the Canada Energy Regulator; rhombic sulphur’s density per PubChem; helium’s leading producers per the USGS Mineral Commodity Summaries 2025; the NGL price relationships per U.S. EIA, the NGL components and ethane rejection per U.S. EIA, and natural gasoline’s place in the NGL barrel per RBN Energy. The crude, natural gasoline and bitumen gravities and densities are converted into one another with the API formula below, and every band is an industry convention, not a standard. The value column is an ordering, not a ratio — every one of these spreads moves with the oil-to-gas price relationship and with local differentials (Table 2).

Crude quality. Two wells can both produce “oil” and sell at prices more than $20 a barrel apart, because crude is a mixture of thousands of molecules carrying contaminants a refiner must remove. Which molecules dominate is decided in the source rock (Section 3.2): the more heat applied, the more the big molecules cracked into small ones, so deep, hot basins make light oil while shallow, cool, biodegraded ones make tar. The industry compresses the result into two numbers.

The first is API gravity, a scale that runs backwards — API = 141.5 ÷ specific gravity − 131.5, so a higher number means a lighter oil. The conventional bands are light above about 30°, medium 22–30°, heavy 10–22° and extra-heavy or bitumen below 10°, and what the number proxies is yield: a light crude is mostly the molecules that become petrol, jet fuel and diesel, a heavy one mostly residue that has to be broken apart. The second is sulphur: low-sulphur crude is sweet, high-sulphur sour, with the line near 0.5%. Sulphur follows the depositional setting — marine carbonate and evaporite source rocks give sour crude, which is why Middle Eastern and Canadian barrels are sour and West African ones sweet — and it is pure cost: US road diesel may carry no more than 15 parts per million (U.S. EIA), so a 3.5%-sulphur crude must shed more than 99.9% of its sulphur first.

Table 2. Crude quality bands, discounts and processing needs

Source: API and sulphur bands per the Wyoming State Geological Survey, with heavy oil’s 22° upper limit per the USGS; the WTI deliverable specification (0.42% sulphur, 37–42° API) per CME Group and the light-sweet series definition per U.S. EIA, 2015; Western Canadian Select’s five-year average quality per CrudeMonitor; the WTI–WCS annual average differentials and their drivers per the Alberta Energy Regulator (ST98), which forecasts roughly US$11/bbl for 2025 after the Trans Mountain expansion opened in May 2024; bitumen gravity, dilbit sulphur and the ~30–35% dilution rate per Oil Sands Magazine, and raw Alberta bitumen’s sulphur of about 5% per Oil Sands Magazine. Per-grade assays for the Middle Eastern, Russian, North Sea and South American streams are carried in the oil guide. The bands are industry conventions, not standards, and the discounts are annual averages, not quotes — a differential moves with refinery outages, pipeline capacity and the sulphur cycle, so re-source it at the time of use. What it costs to produce each of these barrels is a separate question, answered by the cost curve in Table 21.

For a producer, all of this arrives as one number — the quality differential between the benchmark and what its own barrel fetches, the line that opens the netback walk in Table 9. Heavy and sour crude is dearer to move as well as to refine: bitumen cannot meet a pipeline’s specification of roughly 350 centistokes (cSt) and 940 kg/m³, so it is blended with about 30–35% condensate by volume (Oil Sands Magazine) that the producer must buy and ship, and sour production needs sour-service steel, gas sweetening and sulphur recovery before anything is sold (Section 3.4). The takeaway is that quality is an upstream fact: two companies with identical volumes and costs can differ permanently in revenue per barrel, and unlike a cost gap a quality gap never closes — which is why the realised price, not the benchmark, is the line to read.

1.3 Conventional vs unconventional

Not all oil and gas is produced the same way, and the method decides the economics. Conventional hydrocarbons have migrated out of their source rock into a porous, permeable reservoir under an impermeable cap — a discrete trap that a vertical well can drain over decades at a slow, predictable decline. Everything else is unconventional, and the test is one question — will the reservoir give its hydrocarbons up to an ordinary well? Where it will not, it fails for one of three reasons:

• The rock will not flow. Shale and tight sandstone or carbonate hold hydrocarbons in poorly connected pores, so the well creates its own plumbing with horizontal drilling and hydraulic fracturing — water, sand and chemicals pumped at pressure to crack the rock and prop the cracks open (Section 3.3). In coalbed methane the gas is adsorbed onto the coal and released by pumping water off the seam.
• The fluid will not flow. Heavy oil and bitumen often sit in good reservoir rock but are too viscous to move, so the fix is heat, dilution or mining — the subject of the oil sands guide.
• It is not oil yet. Oil shale holds kerogen that has never been cooked into oil and must be heated to around 500°C — which is why the US Green River deposit has never been produced commercially, and why oil shale is not shale oil.

Two edges of the scheme matter: offshore is a location, not a resource class — deepwater geology is conventional, and it gets a row in Table 3 because its economics behave like nothing onshore — and heavy oil straddles the line, a conventional reservoir holding an unconventional fluid.

Table 3. Resource types and how they behave

Source: resource-type taxonomy per U.S. EIA and standard reservoir-engineering practice, with the gas-side categories and the coalbed-methane share of US dry gas per U.S. EIA, 2022, and oil-shale retorting per the U.S. Department of Energy; decline and capital profiles are typical ranges, not measured values, and Section 4.3 puts measured numbers on the decline column. The class column follows common industry usage: a resource is unconventional where the reservoir cannot deliver its hydrocarbons to an ordinary well — because the rock will not flow, the fluid will not flow, or the organic matter is not yet oil — while offshore is a location rather than a class, and heavy oil sits on the boundary. The geology behind these plays, and each basin’s crude quality and extraction method, is mapped in the world oil & gas basins guide.

The short-cycle nature of shale is the defining feature of the modern industry. A conventional megaproject is a ten-year commitment made once; a shale programme is a factory that drills, completes and declines wells continuously, and can be sped up or throttled within months. That flexibility is why US production reacts to price so much faster than it did before 2010 — and why the sector’s capital discipline (Section 7.4) matters so much.

1.4 Units & reading a production number

Upstream forces the reader to hold oil and gas units side by side, because most producers make both. This report follows the industry’s Roman-numeral convention: M = thousand, MM = million. Oil is quoted in barrels (bbl) — 42 US gallons — and flows in bbl/d, Mbbl/d (thousand) and MMbbl/d (million); gas is quoted in cubic feet — Mcf, MMcf and Bcf/d (billion cubic feet per day). The barrel of oil equivalent (boe) ties the two together, so a company can report a single production number (boe/d) and a single reserve number (MMboe) — and that single number conceals what the company actually sells.

Table 4. Units used in this report

Source: unit conventions per U.S. EIA and standard reserve-reporting practice; magnitudes as cited in Sections 5.2 and 6.1. Crude-quality axes (light/heavy, sweet/sour) are set out in Section 1.2 and covered from the market side in the oil guide.

What a boe actually is. The barrel of oil equivalent converts gas to oil on energy — 6 Mcf ≈ 1 boe, roughly the heat in a barrel of crude — and gas-weighted companies run the same ratio backwards in Mcfe. Both carry the same flaw: energy equivalence is not value equivalence. At $70 oil and $3.50 per Mcf, the market’s own exchange rate is 20 Mcf to the barrel, not 6. Canadian filings must print the warning, and Canadian Natural’s wording is worth memorising:

“A BOE is derived by converting six thousand cubic feet (‘Mcf’) of natural gas to one barrel (‘bbl’) of crude oil (6 Mcf:1 bbl). This conversion may be misleading, particularly if used in isolation, since the 6 Mcf:1 bbl ratio is based on an energy equivalency conversion method primarily applicable at the burner tip and does not represent a value equivalency at the wellhead.”

US filers use the same ratio, with the same weakness, and are under no obligation to say so. In 2025 EQT sold about 1.1 million boe/d, 94% of it natural gas, realised at $3.19 per Mcfe (EQT, Q4 2025 results), while Canadian Natural produced a record 1,571 Mboe/d led by upgraded synthetic crude and thermal bitumen, with gas a little over a quarter of the total (Canadian Natural, Q4 2025 report; the boe conversions and shares are this report’s arithmetic at 6 Mcf : 1 boe). Both are “million-boe/d producers” selling almost entirely different molecules — and a bitumen line is even reported net of the diluent blended into it.

Four conventions move the number without moving the rock. Gross versus net: a US filer reports volumes net of royalties, a Canadian one usually before them (Section 2.2). Working interest versus entitlement: under a production-sharing contract cost recovery is paid in barrels, so reported entitlement falls when prices rise (Section 4.6). Produced versus sold: fuel gas, flaring and processing losses make “production” and “sales volumes” different lines. And ethane rejection: when ethane is worth more as heat than as a liquid it stays in the gas stream, moving barrels from the NGL line to the gas line with no change underground.

So four questions strip the ambiguity out of a production headline — which ratio, what product mix, gross or net and produced or sold, and what the company realised per boe. The last does the work the unit refuses to do: EQT’s $3.19 per Mcfe is about $19 per boe, where an oil-weighted producer may realise three or four times that. Comparing producers by volume compares the wrong number; the comparison that means anything is cash per barrel, the subject of Section 1.5.

1.5 How upstream gets paid

Upstream has the simplest and most brutal revenue model in the value chain: revenue equals volume times price. There is no reservation charge like a pipeline’s, no crack spread like a refiner’s, and no contract that fixes the number — a producer sells its oil at a differential to a benchmark such as Brent or WTI and its gas at a differential to a hub such as Henry Hub — at the lease or further downstream (Section 3.4) — and keeps whatever is left after getting the barrel out of the ground.

Three cost measures define that leftover. Lifting cost (or operating cost) is the cash spent to produce a barrel already drilled — labour, power, water handling, maintenance. Finding & development (F&D) cost is the capital spent to add a barrel of new reserves, and it is what a company must beat over time, because every barrel produced must eventually be replaced. The gap between the realised price and those costs is the netback, the true per-barrel profit at the wellhead, and a producer’s breakeven is the price at which netback goes to zero. Section 4.1 takes that cost apart layer by layer.

This is what makes upstream the amplifier of the chain. A pipeline earning a fixed fee is barely moved by the oil price and a refiner is exposed to the spread, but a producer whose costs are fixed sees every dollar of price move drop almost straight to the netback: a $10 rise on a hypothetical $50 barrel with $40 of cost doubles the profit, and a $10 fall erases it (Table 5).

Table 5. How the three value-chain segments earn

Source: revenue-model taxonomy across the value chain from company filings; the per-segment mechanics are developed in each part of this series.

The one modification producers make is hedging — selling forward part of production to lock in a price. It keeps a leveraged producer solvent through a downturn but caps the upside, so how much a producer hedges is a statement about how it wants to be owned.

2. Reserves, resources & inventory

Section 1 described what upstream sells; this section is about the stock it sells from. Those barrels are underground and cannot be counted directly, so the industry grades them by how confident the estimate is — and reading that grading is the single most important analytical skill in the sector, because a “barrel” can mean wildly different things.

2.1 Reserves vs resources

The critical line is between reserves — quantities both discovered and commercially recoverable under the prescribed prices and current technology — and resources, everything less certain: discovered but not yet commercial (contingent) or not yet discovered (prospective). Only reserves drive the accounts; depletion and the standardised measure (Section 7.2) are both calculated on them.

The SPE’s Petroleum Resources Management System (PRMS) grades every volume on two separate axes, and most confusion comes from mixing them. Chance of commerciality — has the company committed to a project that will produce it? — separates reserves from contingent and prospective resources. Range of uncertainty — given that project, how much will it recover? — gives a low, best and high estimate: 1P / 2P / 3P for reserves and 1C / 2C / 3C for contingent resources. For reserves it reads like this:

• Proved (1P) — recoverable with reasonable certainty: at least a 90% chance (P90) of recovering the estimate or more. It splits by development status into proved developed producing (PDP), flowing now; proved developed non-producing (PDNP), drilled but shut in; and proved undeveloped (PUD), booked but needing new wells or a major outlay. PDP adds production history, which makes it the lowest-risk barrel in the stack.
• Probable — the increment to 2P, which is as likely as not to be recovered (P50): the evaluator’s best estimate and the figure most transactions are priced on.
• Possible — the increment to 3P, with at least a 10% chance (P10) of being reached: the high case, rarely filed.

The categories are cumulative — a company reporting “2P of 500 MMboe” already has its proved barrels inside the 500 — and each percentage is a floor rather than a forecast, which is why 1P is the conservative number and 2P the realistic one.

2.2 What it takes to book a barrel

A barrel becomes a reserve when the company clears every gate the rules set. The gates are the same for 1P, 2P and 3P; only the confidence required in the volume changes:

1. Hold the right to produce it — a lease, licence, concession or production-sharing contract, booking only the company’s own entitlement: net of royalties for a US filer, usually “company gross” before royalties for a Canadian one.
2. Prove the accumulation with a well that has established hydrocarbons; under SEC rules proved volumes stop at the lowest known hydrocarbons seen in a well. No well, no reserves — only a prospective resource.
3. Estimate the volume with an accepted method — volumetrics early, decline curves and reservoir simulation later, or any field-tested “reliable technology”.
4. Pass the economic test at the prescribed price — for an SEC filer, the unweighted average of the first-day-of-the-month prices over the prior twelve months; under PRMS and Canada’s NI 51-101, a forecast deck, so the same barrel can be a reserve in one filing and not in another.
5. Commit to developing it — an adopted plan to drill it within five years, the gate that keeps most inventory off the books.
6. Have a qualified evaluator stand behind it — mandatory and independent under NI 51-101 in Canada.
7. Reconcile it every year, in the 10-K (the annual report US companies file with the SEC) or Canada’s Form 51-101F1.

Nor are the gates passed once. When the twelve-month average price collapsed in 2015–16, ExxonMobil’s Canadian oil-sands proved reserves fell from 4.1 billion barrels to 0.4 billion in a single year — the bitumen had not moved; it had failed the price test — before 3.2 billion barrels went back on the books in 2018 (IEEFA) and roughly 3 billion came off again in 2020 (IEEFA).

Three rulebooks set the gates, and they differ exactly where comparisons break. SEC filers must report proved reserves at the historical twelve-month price and rarely disclose probable or possible; Canadian filers under NI 51-101 must report both 1P and 2P at forecast prices, independently evaluated; most other international companies report under PRMS, at forecast prices, with 2P as the headline. Comparing reserves across jurisdictions means knowing which definition, which price deck and which interest — gross or net — is in play.

Table 6. Reserves and resources classification

Source: classification per the SPE Petroleum Resources Management System (PRMS) and the SPE PRMS FAQs; U.S. SEC reserves-reporting rules, codified in Regulation S-X Rule 4-10(a) and Regulation S-K Item 1202; Canada’s NI 51-101 and the CSA’s Staff Notice 51-324 glossary; probability bands are the frameworks’ own definitions. The four inventory rows are this report’s mapping of company-stated locations onto the framework — neither the SEC nor PRMS defines inventory.

2.3 Inventory, booked and unbooked

The bottom four rows of Table 6 place inventory: a company’s stock of identified, undrilled, drillable well locations on acreage it already controls, quoted as a count or, more usefully, as years of drilling. It is not a reserve category. Only the booked slice — locations inside the adopted five-year plan that clear the filing price — has passed the gates, and it is already counted as PUD; everything beyond it is unbooked, at best a contingent resource, and much of it is excluded for timing rather than geology — a company with twenty years of locations can only schedule five.

Companies grade the unbooked stock themselves — Tier 1 or “premium” locations economic at a conservative price, Tier 2 and 3 needing more — and each sets its own hurdle: EOG, for example, defined a “premium” well as one earning at least a 30% direct after-tax return at a flat $40 oil and $2.50 gas. Inventory counts are company-stated, unaudited and often quoted without their price deck, so before comparing two companies’ “fifteen years of inventory”, check how much is booked, the price at which the tiers are cut, the lateral length a “location” assumes, and whether the count is gross or net.

A company with a large 3P number and a small PDP number is selling a story about the future; one with deep PDP and low-cost inventory is selling cash flow today. The gap between the two is where most upstream mistakes are made.

3. From acreage to sales gas

Sections 1 and 2 treat a well as an abstraction. This section walks the real path in the order it happens — the right to drill, the target, the well, what comes out of it, the same at sea, and the recovery work that follows — because almost every cost figure in this guide is decided somewhere along it.

3.1 Leases & tenure

Nothing is drilled until someone holds the legal right to the hydrocarbons underneath, and in oil and gas that right is bought, not staked. Onshore in the United States the mineral estate — the right to produce what lies below — is very often privately owned and severed from the surface, which makes the country close to unique: federal land and the federal offshore Gulf together supply roughly a quarter of US crude, leaving about three-quarters of American oil coming off private and state acreage, leased tract by tract. Almost everywhere else the state owns the subsurface and licenses it out. Figure 3 shows the gates before a legal well and Table 7 the routes to acreage.

Figure 3. The gates before a legal well

Source: sequence per the leasing, pooling and well-permitting frameworks cited through this section; conceptual diagram. Federal production shares above per U.S. EIA (onshore federal lands, 2024) and U.S. EIA (federal offshore Gulf, 2025). The mining equivalent — claim, exploration licence, surface agreement and permit — is walked stage by stage in the claim-to-closure guide.

Table 7. How upstream acreage is acquired in North America

Source: federal onshore terms per the BLM 2024 leasing rule and the 10-year primary term per the Mineral Leasing Act, 30 U.S.C. § 226(e), with the royalty floor returned to 12.5% by the 2025 One Big Beautiful Bill Act per the BLM; offshore rate per BOEM; state royalty per the New Mexico State Land Office; Canadian Crown tenure per Alberta.ca and its tenure overview; Canadian offshore tenure per C-NLOER and CAPP. Bonus levels are negotiated per deal and are not published; federal fees, rentals and royalty floors are set by statute and rule and move with the administration, so re-source them at the time of use.

The US instrument is the oil & gas lease: an upfront bonus per acre, a primary term of commonly three to five years, and a lessor royalty taken off the top before costs — traditionally one-eighth (12.5%), bid up to 20–25% in the best rock, where New Mexico raised the top rate on its premium state leases to 25% in June 2025. The defining clause is held by production (HBP): once a well produces in paying quantities the lease is held for as long as it keeps producing, which makes the primary term a drill-it-or-lose-it clock. A modern two-mile (3.2 km) lateral also crosses many separately owned tracts, which must be pooled by contract or unitized before a permit can issue — broadly by statute in Oklahoma, North Dakota and New Mexico, far more narrowly in Texas — and because the estates are severed, surface access is negotiated separately.

Canada starts from the opposite end: the Crown owns 81% of Alberta’s minerals and issues them at a public offering roughly every two weeks, as tenure that must be validated by drilling and is continued only while the lands are proven productive; the remaining 19% is freehold, leased privately much like a US tract (Table 8).

Table 8. The clauses that decide when acreage is lost

Source: Canadian continuation, validation and non-productivity mechanics per Alberta.ca; US federal rentals per the BLM and the 10-year federal primary term per the Mineral Leasing Act, 30 U.S.C. § 226(e); forced-pooling limits per the Texas Mineral Interest Pooling Act. Private-lease clauses (HBP, Pugh, depth, shut-in) are standard North American drafting rather than statute, so their exact wording is negotiated lease by lease.

The royalty and the taxes stacked on a lease do not attach at the same point. A lessor royalty comes off the top of revenue, a severance or production tax is shared by the royalty and working-interest owners, and operating and transport costs fall on the working interest alone. Table 9 walks one hypothetical barrel through the sequence.

Table 9. Where the royalty and the tax land — one hypothetical barrel

Source: the royalty rates are anchored to the benchmarks in Table 7 — 25% is the top rate on premium state leases per the New Mexico State Land Office and 12.5% is the traditional one-eighth and the current federal floor per the BLM — but the table as a whole is a hypothetical illustration of the mechanism, not a valuation of any company, lease or well: every price, differential, cost and tax figure is an assumption chosen to be plausible, and none is measured. Input marks an assumption of this illustration and Derived the arithmetic of the rows above it; rows are numbered within each band. The Value column is headed Value rather than US$/bbl because the build mixes prices, rates and shares — the unit sits in the line item. Figures are rounded half-up to two decimals.

At a 25% lease with a 2% override, 27% of gross wellhead revenue is gone before a single cost is paid, leaving a netback of US$27.30 against US$33.87 on an otherwise identical one-eighth barrel — which is why the net revenue interest, not the headline working interest, is the number to check, and why inventory is a legal claim on a clock, not just rock: acreage is read for its clock, its contiguity and its royalty burden (Section 7.2). Not every regime charges from the first barrel — Alberta’s Crown royalty waits for payout (Section 4.6) — and how a royalty becomes a tradable security is the subject of the ownership & deal structures guide.

3.2 Where to drill

Long before a rig moves, the operator has built a model of rock it cannot see around the petroleum system — five things that must all be present, in the right order: a source rock buried deep enough to be cooked in the oil window of roughly 60–120 °C; migration of the lighter hydrocarbons upward; a porous reservoir rock; an impermeable seal; and a trap that gathers what arrives. Miss one, or get the timing wrong, and there is nothing to find.

Figure 4. The petroleum system — five things that must line up

Source: standard petroleum-geology framework; conceptual diagram. The elements must also be correctly timed relative to one another, which the linear sequence above cannot show.

Unconventional resources invert the exercise. In shale the source rock is drilled directly, so migration, seal and trap are irrelevant and “exploration” is really appraisal at industrial scale: everyone knows the rock is there, and what is unknown is where it is best — which is why shale could scale so fast once the geological risk was largely removed. For conventional targets the workhorse is seismic reflection — sound sent into the ground and timed on its way back — shot in 2D for reconnaissance, 3D to map a trap’s shape and size, and 4D to watch fluids move as a field produces (Table 10).

Table 10. The exploration toolkit

Source: standard exploration-geoscience practice; the seismic acquisition methods and the conventional-versus-unconventional distinction follow U.S. EIA primers. Which tools a given prospect takes, and in what order, varies with how well the basin is already understood.

None of those tools removes the risk, because only the bit proves it. Westwood’s data put the commercial success rate of high-impact wells at 31% in 2025, and at 27% across 2020–24 against 21% in 2010–14 — better imaging has helped — but in ultra-deepwater beyond 2,500 m it was 3% in 2020–24, one success from 35 wells, and only 64 high-impact wells were completed in 2025, the lowest on record, as capital went into buying production instead (Section 7.4). Offshore, seismic is cheap next to a single well, so operators shoot as much as they can and drill as little; onshore, shale wells are cheap enough to be the exploration tool.

3.3 Drilling & completion

Drilling starts with the spud and proceeds in telescoping sections, each lined with steel casing cemented to the rock before the next begins. The surface casing, set below the deepest usable groundwater, is the well’s primary environmental barrier, and only once it is set is the blowout preventer (BOP) — hydraulic rams that can close the well in seconds — bolted on. On a shale well the bit then turns from vertical through a curve, lands inside a target layer that may be only 50–150 ft (15–45 m) thick, and drills the lateral along it to total depth (TD). Weighted drilling mud holds back formation pressure and lifts the cuttings, and sensors just behind the bit let the driller geosteer — adjust in near-real time to keep a two-mile (3.2 km) lateral inside the pay.

Wells are drilled in batches from a single pad, spreading the pad, roads and mobilisation across every well, and they have grown fast: the average new Permian lateral went from 6,149 ft (1,874 m) in 2015 to 10,867 ft (3,312 m) in 2025, and “super-laterals” beyond 15,000 ft (4,570 m) were 15% of 2025 completions — the quiet reason US production kept climbing on a falling rig count. A modern Permian horizontal takes roughly two to three weeks from spud to rig release and a few months to first production (Table 11), which is what “short-cycle” actually means.

Table 11. The onshore shale well, spud to first flow

Source: Permian lateral lengths per U.S. EIA, 2025; DUC counts per U.S. EIA, October 2022 (the 2020 record) and the EIA Drilling Productivity Report (April 2024); drilling cycle times of roughly 15–20 days per Permian well in 2025, with leading operators near 13 days spud-to-rig-release, per Oil Gas Leads. The per-stage split is the author’s indicative allocation of that cycle, not a measured statistic — operators publish the total, rarely the breakdown — and every band describes one well type: a modern onshore US shale horizontal on a multi-well pad. Angles, section lengths and target thicknesses are likewise typical ranges that vary widely by basin and well design. A fractured well uses 1.5–16 million gallons of water per the USGS, in a mix of roughly 85% water, 14% proppant and under 1% additives disclosed well by well on FracFocus, with proppant and fluid volumes per well broadly flat since 2017 per U.S. EIA, 2021.

Figure 5. From spud to sales gas

Source: drilling and completion sequence per Table 11; gas treatment stages per U.S. EIA; conceptual diagram.

Offshore, the deepwater subsea well. The physics are the same, but a floating semisubmersible or drillship replaces the land rig, the BOP moves to the seafloor on a marine riser that can be more than two miles long, and the reservoir is usually completed with sand control rather than fracturing, then tied through a subsea tree into a seabed manifold. A deepwater well commonly takes one to three months or more of rig time at a day rate near $500,000 (Table 12); ExxonMobil’s $9bn Payara project in Guyana was sanctioned in 2020 on up to 41 wells — 20 producers and 21 injectors that push water or gas back to hold reservoir pressure — to fill a 220,000 bbl/d floating production, storage and offloading vessel (FPSO, Section 3.5) that started up in late 2023 (ExxonMobil, Hess). Onshore a well is a production-line item; offshore every well is a project.

A cemented steel pipe produces nothing until it is completed. On a shale well the dominant method is plug and perf — isolate a stage, perforate it with shaped charges, then hydraulically fracture it with water and proppant (sand) — repeated often 50–100 times along the lateral, at roughly half the well’s cost, and completion design, more than geology alone, separates a good well from an average one in the same rock.

Two consequences follow the fluid. The water has to be sourced, moved and disposed of — first as flowback, then for decades as produced water (Section 3.4). And fractures do not respect a drilling schedule: a new “child” well fractured beside an older “parent” can hit its depleted rock and damage both, which is why a pad is fractured in one campaign and why well spacing matters (Section 4.4).

Between the rig and the frac crew, a shale well waits as a drilled but uncompleted well (DUC) — cased to total depth with no opening in the pipe, so it can neither produce nor leak. US DUCs peaked above 8,800 in mid-2020 and stood near 4,500 in April 2024. The wait is partly logistics, but it is also the cheapest optionality in the business: the drilling is sunk, so a gas producer can park wells through a weak strip — the run of futures prices for the coming months — and complete them when it improves, which makes the DUC count the fastest supply response US shale has. When pumping stops, the well flows back its injected water, then hydrocarbons, and its steep decline begins (Section 4.3).

The two kinds of well are also two different cheque sizes, four orders of magnitude apart (Table 12). Capital scale decides who can attempt a project — a $10bn hub is open to a supermajor, a national oil company or a consortium and nobody else (Section 6.2) — and it decides reversibility: a shale programme can be slowed within a quarter, a sanctioned deepwater project cannot, the asymmetry that the full-cycle breakeven (the price that justifies building) and the half-cycle breakeven (the price that keeps an existing facility running) measure in Section 4.2.

Table 12. The capital scale of upstream, one well to one project

Source: shale drilling and completion costs per Permian Resources, Q3 2025 results (~US$725 per lateral foot), and Permian Resources, Q4 2025 results (~US$700 per foot, and ~8% lower per foot guided for 2026); drillship day rates per the Transocean fleet status report, July 2025 — the primary dayrates on its ultra-deepwater drillships in the US Gulf, one of which runs higher; the Whiptail contract value, stated as between US$750 million and US$1.5 billion, per World Oil, April 2024; Yellowtail’s sanction value per Offshore Technology and its start-up per ExxonMobil, August 2025; Blackrod Phase 1 capital and first oil per International Petroleum Corporation, June 2026. Capital costs move with the service cycle, steel prices and scope — every figure here is a dated benchmark to re-source at the time of use, not a standing rate. The named projects are factual examples of scale, not recommendations, and the per-flowing-barrel figure is derived from the two disclosed numbers beside it. Deepwater well costs also carry the stricter US well-control requirements introduced after the 2010 Deepwater Horizon disaster — dual shear rams and real-time monitoring — per BSEE.

3.4 From wellhead to sales gas

What leaves a producing well is a mixture of gas, liquid hydrocarbons, water and often hydrogen sulphide and carbon dioxide, and turning it into something saleable is the last upstream step. Oil wells and gas wells part company here. As reservoir pressure fades, an oil well needs artificial lift — a rod pump, an electric submersible pump or gas lift — while a gas well usually flows on its own for far longer. And oil can sit in a tank, so a new oil well can leave by truck on day one, whereas a gas well without a pipeline connection produces nothing at all — the asymmetry behind shut-in gas wells, the flaring of associated gas where gathering runs short, and a gas producer’s need to think about takeaway before drilling.

At surface the stream passes the wellhead and the valve stack of the Christmas tree, where a choke trades rate today against recovery later, and enters a separator that drops the pressure so gravity can sort it: gas rises, oil floats on brine. The oil goes to a tank battery and leaves by truck or pipeline; the water goes to a saltwater disposal well or is recycled, a cost that often exceeds the oil in volume; and the gas enters a gathering system, is compressed, and reaches a processing plant that dehydrates it, strips out hydrogen sulphide and carbon dioxide, and separates methane from the heavier natural gas liquids (NGLs), which are fractionated and sold separately.

Every outlet is metered: gas in Mcf and Btu, oil through a LACT unit (lease automatic custody transfer) that also checks basic sediment and water and marks the legal point where the barrel changes hands — the meter the royalty cheque in Table 9 is built on — with periodic well tests allocating commingled production back to each well.

Where upstream ends is a meter, not a place. Custody always passes there; ownership passes too when the barrel is sold at the lease, while a producer that sells further downstream keeps title and pays someone to move, treat or store its molecules — and everything past the processing plant is Part 2. Who owns the steel is a separate, commercial question (Table 13): gathering lines, compressors and even the plant may belong to the producer, an affiliate or a third party, which is why gathering and processing show up as a deduction in the netback (Table 9) rather than as a business the producer runs — and why a well with no gathering line or takeaway flares, sells at a punishing differential, or does not flow.

Table 13. From wellhead to pipeline — who owns what

Source: the physical sequence per U.S. EIA and standard field practice; ownership is a commercial question, not a technical one — the column shows the common case and varies by basin, contract and vintage, with Canadian and legacy US fields more often producer-owned. Everything below the boundary is covered in the midstream guide.

Who buys the barrel. The meter settles custody; the sale contract settles who pays to get the barrel to market, and producers take one of three routes (Table 14). Most small and mid-size onshore oil producers sell at the lease: a crude marketer takes ownership where the barrel is measured on its way off the property, which is where the U.S. EIA records a crude oil first purchase (U.S. EIA), and pays the benchmark less a differential for quality, energy content, transport and location (Netherland, Sewell & Associates). The buyer arranges the trucks and pipe; the producer pays for them through the differential.

Larger producers keep title to a hub. They dedicate acreage or commit minimum volumes to gathering systems before the wells are drilled, reserve firm transportation on long-haul pipelines, and sell where the price is set: EQT reported access to about 4.9 Bcf/d of firm pipeline takeaway capacity and 1.1 Bcf/d of firm processing capacity (EQT, 2024 Form 10-K, p.15). Canadian producers did the same on the Trans Mountain expansion, where about 80% of capacity is reserved for committed shippers on long-term take-or-pay contracts (Canada Energy Regulator, 2025) — Canadian Natural, Cenovus, Imperial and Suncor among them (Trans Mountain, 2013). Integrated majors sell through their own trading arms, and offshore fields sell cargoes: each partner lifts its share, as Guyana’s government sold its first Liza entitlement to Shell in 2020 while ExxonMobil and Hess lifted their own (S&P Global, February 2020).

Gas also passes through a processing contract — fee-based, percent-of-proceeds or keep-whole — explained in the midstream guide, “How midstream gets paid”. Either way, the producer pays for market access, as fees or as a wider differential, so the realised price after transport deductions is the comparable number — and firm commitments are fixed costs that stay when production falls.

Table 14. How a producer sells its barrel — three routes and the gas contract

Source: control-transfer points — wellhead, plant inlet or tailgate, or a third-party delivery point — per EY, “How the revenue recognition standard affects upstream oil and gas entities”, January 2025; the lease as the point of a crude oil first purchase per U.S. EIA; firm takeaway and processing capacity per EQT, 2024 Form 10-K; committed shippers per the Canada Energy Regulator, 2025, and Trans Mountain, 2013; FOB cargo sales per S&P Global, 2020; processing contract types per the midstream guide, “How midstream gets paid”. Routes are typical patterns, not rules — one company can sell some barrels at the lease and hold firm capacity for others, and every contract term is negotiated.

What comes out is more than oil and gas (Table 15). The valuable co-products are on the gas side: NGLs price off crude, export and petrochemical demand rather than Henry Hub, which is why a liquids-rich well outearns a dry-gas well at the same Mcf rate (U.S. EIA); sweetening sour gas yields sulphur, of which the United States recovered about 7.6 million tons in 2025 across refineries and gas plants; and a few nitrogen-rich streams yield helium, about 2.9 Bcf of it sold in 2024. The cost streams are larger by volume, and produced water dwarfs the rest: the Permian lifts three to four barrels of water for every barrel of oil, some 22 million barrels of water a day in 2025, and in parts of the basin water, not rock, has become the binding constraint on growth. A lease holds only oil, condensate and that water in tanks; everything else is moving rather than stored, and hydrogen sulphide is converted the moment it is removed.

Table 15. What else comes out of a well

Source: recovered sulphur per the USGS Mineral Commodity Summaries 2026 and helium per the USGS Mineral Commodity Summaries 2025; Permian water cuts and volumes per The American Oil & Gas Reporter; gas-processing streams per U.S. EIA. Which streams a given well produces depends entirely on its fluid and its reservoir — most wells produce only some of these.

3.5 Offshore production systems

At sea, the separator, tanks and meters that sit on a cheap gravel pad onshore must stand on a structure in open water — usually the single largest cheque in the development — and which structure gets built depends on water depth, whether a pipeline already exists, and field size (Table 16).

In shallow water the facility stands on the seabed — a steel-jacket fixed platform, the cheapest answer wherever it can reach, which is why the Arabian Gulf and the North Sea shelf are full of them. Past roughly 500 m it floats: a tension leg platform or spar suppresses heave enough to put the well valves on deck; a semi-submersible production unit processes but does not store; and an FPSO — a ship-shaped floating production, storage and offloading vessel — does everything, holding its oil in the hull for a shuttle tanker. Shell’s Turritella, in 2,900 m of water, was the deepest production unit ever installed.

Storage decides the map. The Gulf of Mexico has a dense pipeline grid, so its deepwater fields push oil into pipe; Guyana, Brazil and West Africa have none, so their fields store and export by tanker, which is why a Guyanese development is known by the name of its vessel — usually leased from a contractor, so it appears in the producer’s accounts as a lease rather than as capital spending.

The cheapest offshore barrel comes from not building a facility at all. A subsea tieback runs seabed wells to a host with spare capacity, so a discovery’s value depends as much on its distance from existing steel as on its size — the same barrels are worth far more 15 km from a host than 80 km from anything — within the limits of flow assurance: the waxes and hydrates that form as oil cools in a long flowline.

Table 16. Offshore production systems, by water depth

Source: structure types and their deepest installed examples as of 2005 per NOAA Ocean Explorer — newer installations of several types sit deeper, so the reach column is a floor rather than a limit; the Turritella FPSO’s 2,900 m depth, disconnectable turret and lease terms per SBM Offshore, 2016; tieback distances and their capital and schedule advantage per Offshore Engineer, 2024. Depth reaches are practical ranges rather than limits — each type has been pushed deeper than its typical use — and the deepest-example figures are dated records — NOAA’s to 2005, Turritella’s from 2016 — not the norm.

For an investor the structure is where offshore’s economics live: sunk capital with a twenty-to-thirty-year life, whose spare capacity is itself an asset and whose decommissioning liability grows with age, and which makes the marginal barrel cheap once built — the half-cycle economics of Section 4.2.

3.6 Recovery factors & enhanced oil recovery

Every number so far describes the oil that comes out; the one that decides how much that is — and that almost no company headlines — is the recovery factor, the share of the oil originally in place that will ever be produced (Table 17). Primary recovery, on the reservoir’s own energy and then artificial lift, typically produces about 10% of it; secondary recovery — water or gas injected through injector wells to hold pressure up — takes cumulative recovery to roughly 20–40%; and enhanced oil recovery (EOR) can lift the total to 30–60% or more, above all by injecting carbon dioxide, which is why the Permian has a CO₂ pipeline network, or steam, in California’s heavy-oil fields and, at industrial scale, Alberta’s steam-assisted gravity drainage (SAGD) projects (see the oil sands guide).

Table 17. Recovery factors and how they are raised

Source: primary, secondary and tertiary recovery percentages and the US EOR mix by method per the U.S. Department of Energy; shale oil and gas recovery efficiencies per the U.S. EIA and ARI resource-assessment methodology, 2013; SAGD recovery per Oil Sands Magazine and the mining plant-recovery requirement — 90% by weight for ore of 11% bitumen or richer — per the Alberta Energy Regulator, Directive 082 (2016). Recovery factors are reservoir-specific and these are typical ranges, not guarantees; the thermal row’s own recovery is the oil-sands figure in the oil sands guide.

For an investor, shale’s economics are not a recovery story: a tight-oil well produces perhaps 3–6% of the oil beneath it, and the play works because the in-place volumes are enormous and the wells cheap, not because extraction is efficient — so every point of recovery unlocked by refracturing, tighter spacing or gas injection would add reserves without a new lease. In a mature field, recovery work is where the remaining barrels are, and it shows up in filings as a reserve revision once the project is committed (Section 2.1): the rock did not change, the plan did.

4. The economics of the marginal barrel

Section 3 followed the barrel out of the ground. This section prices it: what it costs, the price at which it is worth producing, how fast it declines and what that forces a company to spend, and the two outside hands — OPEC+ and the state’s fiscal terms — that set its terms.

4.1 What a barrel costs

A breakeven is a sum, and the most useful thing an investor can do with one is take it apart. Operating costs are the cash spent to keep existing wells flowing — what the half-cycle breakeven (Section 4.2) measures. Capital costs are the money spent to find, drill and build, part merely to replace what decline took and the rest to grow — what the full-cycle breakeven adds. And claims on revenue — royalties, production taxes and midstream fees — come off before the producer sees a netback at all. Table 18 lays out the stack in the order a barrel meets it, with the label each layer carries in a US and a Canadian filing.

Table 18. The upstream cost stack, and what the filings call each layer

Source: filing labels per the 2025 reports of Diamondback Energy (FY2025 results), EQT (Q4 2025 results) and Cenovus Energy (2025 annual MD&A); the capitalised-versus-expensed treatment of exploration spend per U.S. SEC Regulation S-X Rule 4-10. Filing labels vary between companies — the quoted ones are common forms, not mandated wording — and sustaining and growth capital have no accounting definition at all, so each company draws its own line. The per-barrel figures for each layer are in Table 19.

Lifting cost (Section 1.5) is therefore a bundle, and on an onshore oil well its fastest-growing item is produced water: at the US$0.84 per barrel of produced water that Aris Water Solutions earned for handling it in 2024 (Aris Water Solutions, FY2024), the Permian’s three to four barrels of water per barrel of oil cost roughly US$2.50–3.40 per barrel of oil on this report’s arithmetic. For a gas producer the lifting is almost free and the pipe is not: gathering, transmission and processing were three-fifths of EQT’s 2025 operating cost (Table 19), which is why the ranking of the lowest-cost US gas producers comes down almost entirely to who owns or negotiates their midstream.

Capital is split by purpose, not by accounting: sustaining capital holds production flat and growth capital is everything above it, both company-defined and unaudited — so test them against the decline rate (Section 4.3) and DD&A, the non-cash charge that expenses past capital as its barrels are produced. For a shale producer DD&A dwarfs the cash cost — Diamondback’s US$14.99/boe in 2025 against US$10.23/boe of cash operating cost — while Saudi Aramco reported just US$3.51/boe of lifting cost and US$8.0/boe of upstream capital expenditure per barrel (Saudi Aramco, Annual Report 2025, p.49). The accounting method moves the numbers too: successful efforts expenses seismic and dry holes as they happen and full cost capitalises them, so identical wells can print different earnings.

Table 19 sets company-reported 2025 operating costs out from the lightest barrel to the heaviest, and Figure 6 ranks the ones printed in US dollars.

Table 19. Operating cost by type of barrel, 2025

Source: Canadian Natural’s operating costs by product line per Canadian Natural, Q4 2025 report, which prints US-dollar equivalents; Cenovus per-asset fuel and non-fuel costs and oil sands transportation and blending per Cenovus Energy, 2025 annual MD&A; Syncrude cash operating cost per Suncor Energy, Annual Report 2025; Petrobras lifting costs per Petrobras, Performance Report 4Q25, “E&P results” table, p.16; Saudi Aramco per its Annual Report 2025, p.49; Diamondback per Diamondback Energy, FY2025 results, and EQT per EQT, Q4 2025 results (gathering US$0.08, transmission US$0.42 and processing US$0.14 per Mcfe). Every company defines operating or lifting cost its own way — some include transport or energy, Petrobras excludes leases and production taxes, and the Canadian lines are in Canadian dollars where no US-dollar figure is printed — so read the column as an ordering, not a like-for-like ranking.

Figure 6. Operating cost per barrel by type of barrel, 2025

Chart source: data in Table 19, limited to the figures the companies print in US dollars — Saudi Aramco’s lifting cost, Petrobras’s pre-salt and post-salt lifting costs, Diamondback’s lease operating expense, and Canadian Natural’s Pelican Lake, North America light crude and NGLs, primary heavy and mining-and-upgrading operating costs — plus Canadian Natural’s thermal in situ operating cost of US$7.93/bbl from the same Q4 2025 report, shown in place of Cenovus’s Canadian-dollar figures. Bars scale to the highest figure in the set (US$17.25). Company definitions differ, so the chart shows an ordering, not a like-for-like cost.

Five patterns come out of it. Scale beats quality in lifting cost: Petrobras lifts its pre-salt barrels for a quarter of what its mature post-salt fields cost, because a platform’s costs are largely fixed. Heavy is not automatically dear to lift — Canadian Natural’s Pelican Lake heavy oil cost less than its light crude — but it is dear to move: Cenovus’s transport and blending came to about four-fifths of its oil sands operating cost. Thermal barrels are energy barrels, burning gas for steam and so structurally short natural gas. Upgrading buys back the discount, at the highest operating cost on Canadian Natural’s North American slate. And sour costs money twice — in sour-service steel, gas sweetening and sulphur recovery hidden inside operating cost (SPE PetroWiki), and in the sour discount on the price.

How producers bring costs down. The levers sort into structural gains that stay when the cycle turns — longer laterals (Section 3.3), multi-well pads, simultaneous fracturing, the scale that contiguous acreage buys and that the Permian merger synergies of Section 7.4 claim, and fixed plants run harder — and cyclical ones: rigs, crews and pressure pumping get cheaper in a bust and dearer in a boom, and a unit cost that fell because a contractor cut its rate is not a gain the producer owns.

Four checks turn per-unit figures into a comparison (Table 20). Put every company on one basis, because a US unit production cost excludes severance taxes and a company-defined lifting cost can leave out whole layers — Petrobras’s Brazil figure more than triples once leases and production taxes go back in. Watch the 6:1 denominator, which flatters gas producers (Section 1.4). Find the capital hiding off the operating line, where workovers, overhead and a full-cost company’s dry holes can sit. And tie cost to reserves: divide the costs incurred plus the change in future development costs by the reserves added, and set that F&D cost against the netback — the recycle ratio of Section 7.2.

Table 20. Where the cost lines sit in a filing, and the trap in each

Source: unit production cost per U.S. SEC Regulation S-K Item 1204; full cost and successful efforts per U.S. SEC Regulation S-X Rule 4-10; the costs-incurred categories per FASB Statement No. 69, now codified in ASC 932; the Canadian per-unit netback disclosure per Form 51-101F1, and the 2015 repeal of the prescribed finding-and-development formula with its replacement disclosure rule for oil and gas metrics per the OSC; Petrobras as in Table 19; Cenovus’s fuel-cost driver per its 2025 annual MD&A. Labels are common forms rather than mandated wording — find the line by what it measures, not by its name.

4.2 The cost curve & breakevens

Because upstream is a price-taker, the question that decides everything is at what price a barrel is worth producing. Ranked from cheapest to dearest, the world’s supply forms a cost curve, and the price is set where it meets demand: producers below it make money, those above it do not drill. That is also how scarcity shows up in this segment — not as a basis blowout (a pipeline filling up, Part 2) or a widening crack spread (Part 3), but as a higher price and, behind it, a higher marginal breakeven as the cheap rock is used up.

The spread along the curve is enormous. Saudi and Iraqi conventional oil lifts for a few dollars a barrel, while the Dallas Fed’s first-quarter 2026 survey found US producers needed about $66/bbl WTI on average to profitably drill a new well — $67 in the Permian, $62–70 across the other regions, and $59 for large producers — which is why oil sliding toward $60 slows US drilling quickly while barely touching Gulf output.

There is no such thing as the breakeven. The full-cycle breakeven carries the upfront capital and its required return and decides what gets sanctioned; the half-cycle breakeven covers only keeping an existing facility running — operating cost, diluent, transport and the quality differential — and decides what keeps flowing in a bust. The same survey gives both for US shale: about $66/bbl to drill a new well but only $43/bbl to cover operating expenses on existing wells, so a price between the two stops the drilling without stopping the production.

Table 21. Approximate breakevens across the cost curve, full-cycle vs half-cycle

Source: US shale breakevens per the Dallas Fed Energy Survey, Q1 2026 (fielded 11–19 March 2026) — $66/bbl on average to profitably drill a new well ($67 in the Permian, $62–70 in the other regions, $59 for large firms), $43/bbl to cover operating expenses on existing wells ($32 for large firms, those producing 10,000 bbl/d or more); oil sands half-cycle per S&P Global Commodity Insights (2025 range and average, WTI-equivalent, including diluent, transport to Cushing and the heavy–light differential) and greenfield SAGD per Enverus Intelligence Research, April 2025. Middle East and offshore full-cycle figures per Rystad Energy, 2024 — average Brent breakevens for new projects: onshore Middle East US$27/bbl, offshore shelf US$37/bbl and deepwater US$43/bbl. That is a different method and benchmark from the survey’s, and every average hides a wide spread by asset, so treat the table as an ordering rather than precise figures. Breakevens move with cost inflation and efficiency and should be re-sourced when used.

Figure 7. Full-cycle vs half-cycle breakeven, oil sands and US shale

Chart source: data in Table 21; the four rows are the ones for which both cycles are published on a comparable basis. The large-producer pair plots the survey’s figures for firms producing 10,000 bbl/d or more ($59/bbl new, $32/bbl existing), and where Table 21 gives a range the bar plots a single representative figure — the oil sands half-cycle bar plots the $27/bbl average of the $18–45 range. Bar lengths are honest against one another; the printed values are the figures cited in Table 21.

The gap between the two columns says how a producer behaves when the price falls. Oil sands has the widest: top-tier greenfield SAGD needs a stable $80+/bbl WTI to be sanctioned and the last new mine, Fort Hills, took its investment decision in 2013, but a built plant is one of the cheapest barrels on earth to keep flowing and barely declines, so growth now comes from debottlenecking. Shale looks similar on paper, but its base falls 30–40% a year (Section 4.3), so the $43 keeps only the wells it already has. The difference is sustaining capital — a small charge on a barrel that barely declines, the whole drilling programme for one that does — which is the subject of Section 4.4.

The cost curve also explains the sector’s self-correction: when the price falls below the marginal breakeven, drilling stops there first and supply growth slows — the same self-correction a refinery closure brings downstream, but far faster, because a shale programme can be throttled in months, not years.

4.3 Decline rates by well and field

What makes upstream unlike any other business is decline: every producing well loses rate as it drains its reservoir, and what that forces a company to spend is the subject of Section 4.4.

Decline is not one number. The IEA’s 2025 study of some 15,000 fields puts the average post-peak decline of a conventional oil field at 5.6% a year and a gas field at 6.8%, but size matters most — supergiants 2.7%, small fields 11.6% — then location — onshore 4.2%, deepwater 10.3% — and region: Middle East fields 1.8%, European ones 9.7% (IEA, 2025; Figure 8). Those are observed rates, with infill wells and waterfloods included; the natural decline beneath them is about three points higher, and if investment stopped everywhere world oil output would fall by about 8%, or 5.5 MMbbl/d, every year — which is why nearly 90% of upstream investment since 2019 has gone into offsetting decline rather than growth.

Shale does not follow the field pattern at all. A shale well starts at its peak and produces around 80% of its lifetime output within two years, against under 10% for a conventional well — a figure that depends on the horizon, as the two paragraphs below set out; a shale basin that stopped drilling would lose more than 35% of its output in twelve months (IEA, 2025), and Permian wells often lose 65–85% of their rate in year one (JPT, 2019). Between December 2023 and December 2024, existing US onshore wells lost 4.3 MMbbl/d of oil and new wells added 4.4 MMbbl/d, just enough to hold output near 11.2 MMbbl/d (U.S. EIA, 2025) — a base decline of close to 40% a year on this report’s arithmetic.

The steep part does not last: engineers fit an Arps decline curve and switch to a constant terminal decline once the rate flattens — most assume 4–7% a year, and a study of about 14,000 US wells found 6.4% in years 10–30 (Payne Institute, 2025). Other assets have their own shapes (Table 22): regulators model a large deepwater project as a four-year plateau followed by 12% a year (BOEM, 2022), coalbed methane rises for three to five years while the seam dewaters, and an oil-sands mine barely declines until its reserves run out.

In barrels, “a well” spans four orders of magnitude (Table 22a). Of the 918,481 wells producing in the United States in 2024, 78% were strippers making 15 boe/d or less (U.S. EIA), which is why the average American well makes about 36 boe/d on this report’s arithmetic, against 18,682 bbl/d for the average Brazilian pre-salt well (ANP, June 2026). A modern two-mile Permian horizontal sits between them: it starts near 1,100 boe/d (derived) and, on the modelled curve of Table 22a, gives up roughly 220–240 Mboe in its first year and perhaps 0.7–1.1 MMboe over a thirty-year life. Set that against the US$7–8m it costs (Table 12) and the well is about US$7,000 per flowing barrel in month one — but it is down to some 330 boe/d a year later, or US$21,000–24,000 per flowing barrel, which is roughly what a new oil-sands phase costs on day one and keeps costing for decades. Shale buys its barrels cheaply and must buy them again; a conventional barrel is dear once and then nearly free.

The 80% needs a horizon. The IEA’s figure describes how front-loaded the curve is, not a thirty-year total. Run an Arps curve fitted to the decline rates above out to the 30-year economic life the EIA assumes and the first two years are nearer 30–45% of ultimate recovery, with 80% arriving around year twelve. Both describe the same well: the tail is long, thin and worth almost nothing once discounted. Only the front of the curve carries value, which is why a shale producer is judged on inventory and F&D cost rather than on reserve life (Sections 4.4 and 7.2).

A company’s decline is the weighted sum of its assets. Canadian Natural, with about 55% of its output from long-life, low-decline assets, reports a corporate decline of about 12% (Canadian Natural, April 2026); Peyto, an Alberta gas producer, expects 26–28% in 2026 (Peyto). For an investor it is the first number to find, because it sets how much of next year’s capital budget is already spoken for.

Table 22. How production declines, by type of well and field

Source: observed and natural post-peak decline rates by field size, location and region, the no-investment scenario and the shale output profile per the IEA, The Implications of Oil and Gas Field Decline Rates, September 2025, oil and gas field tables; Permian first-year well declines per IHS Markit data reported by JPT, December 2019; the deepwater profile per BOEM, OCS Report BOEM 2022-022, a forecasting assumption for projects above 200 MMboe; terminal decline per the Payne Institute, June 2025; coalbed methane per SPE PetroWiki; oil-sands profiles per the Canada Energy Regulator modelling methods. The IEA’s observed rates are compound annual declines since each field’s peak, weighted by cumulative production, and include the effect of continued investment; the BOEM and CER profiles are modelling assumptions, not measurements. The last column is the author’s characterisation of how each asset type is typically sustained, not a measured value.

Table 22a. One well in barrels — rate, lifetime output and life, by type of well

Source: US well counts and the 15 boe/d stripper threshold per U.S. EIA, 2024 data year — 918,481 producing wells, of which 78% made 15 boe/d or less in 2023–24 — with the definition per U.S. EIA; Permian new-well productivity of 960 boe/d across the initial three months of 2021 wells per U.S. EIA, 2022; play-average EURs for remaining unproved, undrilled locations per U.S. EIA (AEO2023 oil and gas supply module, as of 1 January 2021 — Wolfcamp 0.282 MMbbl of oil and 0.835 Bcf of gas per well, Haynesville 9.2–10.5 Bcf, Marcellus interior 3.9 Bcf) and the 30-year economic life per U.S. EIA (AEO2026); Haynesville initial rates and lateral lengths per Comstock Resources, Q2 2026, and the Marcellus 60-day rate per Antero Resources, Q1 2026; Guyana well counts and FPSO capacities per ExxonMobil (Liza Phase 1, Liza Phase 2, Payara and Yellowtail); Brazilian pre-salt volumes, well counts and the highest-producing wells per ANP, June 2026 bulletin; the deepwater EUR per Wood Mackenzie, February 2024; Haradh per-well target rates per JPT, 2015, and the ANDR-1 well’s history per Aramco, February 2025; Firebag and MacKay River well-pair counts and 2024 production per Suncor Energy (2024 Annual Information Form), Blackrod’s 197 planned well pairs and 218 MMbbl of phase-one 2P reserves per Oil Sands Magazine on IPC’s 2023 sanction disclosure with the 25-year plateau per International Petroleum Corporation, June 2026, and the ~10-year life of a pair per Natural Resources Canada, 2024. The rows order the types of well; they do not benchmark them against one another. Rates mix first-month, first-three-month, 60-day and annual-average bases, oil-only and boe bases, and design capacity against measured output, so no cell should be divided into another. Cells marked derived are this report’s arithmetic on the two cited inputs: the Permian first-full-month rate divides the 433,000 bbl/d and 780 MMcf/d that new Permian wells produced in their first full month of July 2024 (U.S. EIA) by the ~500 Permian completions a month — ~6,000 a year since 2022 (U.S. EIA, 2026); the shale-gas year-one figure applies Rystad Energy’s median of 0.574 Bcf per 1,000 ft of lateral for 2022-vintage Louisiana wells, reported by OilPrice (November 2025), to Comstock’s 11,835 ft; the Guyana rate divides 810,000 bbl/d of nameplate capacity by 69 producing wells across the four developments; and the SAGD rates and recovery divide each project’s production and reserves by its well-pair count. No operator in the Permian or the Haynesville now publishes a type-curve EUR, which is why the shale rows carry agency play-averages instead — and those describe all the rock a play has left rather than the rock an operator drills first, so they sit below what a modern long-lateral well recovers. The 0.7–1.1 MMboe range is modelled, not measured: an Arps hyperbolic (nominal initial decline 2.0–3.5 a year, b of 1.0–1.2) scaled to reproduce 960 boe/d across three months, switched to the 6.4% terminal decline above and run to 30 years. The average-US-well figure divides the 2025 production totals of Section 1 by the 2024 well count, the latest published.

Figure 8. Observed post-peak decline of conventional oil fields

Chart source: the IEA, September 2025 — the size and location rows from its conventional-oil table and the regional rows and world average from its executive summary. Each bar is a different cut of the same field database, so the rows overlap (a Middle Eastern field can also be onshore and supergiant) and should be read as a ranking of field characteristics, not of separate groups. Bars scale to the highest value (11.6%). Tight oil and shale are excluded, because their decline is a single well’s steep early fall rather than a field’s post-peak rate.

4.4 Reinvestment & the treadmill

Decline turns upstream into a treadmill: a company must spend continuously just to hold production flat, and only spending above that level grows it — which is why a shale producer’s half-cycle breakeven understates what it actually costs to stand still. Of the 11.3 MMbbl/d the Lower 48 produced in 2025 outside the Gulf of America — 83% of the record US total — wells drilled that year contributed 2.9 MMbbl/d and the entire legacy base 8.3 MMbbl/d, itself falling without reinvestment (U.S. EIA). Reserve replacement — adding at least as many barrels as are produced — is therefore the survival test, and F&D cost (Section 1.5) decides whether it creates or destroys value: a company that pays more to replace a barrel than it earns producing it is liquidating in slow motion, however healthy its production looks.

The same treadmill runs through inventory (Section 2.3). Holding it is close to free once acreage is held by production, and expensive while a primary term is still running, because the clock puts wells on the lease’s schedule rather than the price’s (Section 3.1). But inventory is not maintained — it is consumed: every well deletes a location, and there are only three ways to add one — buy it, downspace at the risk of parent–child interference, or step out into a new bench. That is what the consolidation wave of Section 7.4 bought, and because operators drill their best rock first, inventory years can look stable for a decade while the quality of what remains erodes.

That is also the hinge of the investment debate — whether shale’s steep decline is a race that must eventually exhaust its best rock, or a flexibility that lets the industry respond to price fast and avoids the stranded-capital risk of decade-long megaprojects — and which reading dominates depends entirely on a specific company’s inventory depth and cost position.

4.5 OPEC+ and the coordinating force

Upstream has no regulator setting the terms on which the marginal barrel earns. What it has instead is OPEC+ — the Organization of the Petroleum Exporting Countries plus allies led by Russia — and because a dozen states control most of the world’s low-cost reserves and spare capacity, their quota decisions are the most important recurring event in the oil market.

The mechanism is spare capacity: output a country could add but deliberately holds back. Saudi Arabia carries most of it — the gap between roughly 9–10 MMbbl/d of output and ~12 MMbbl/d of maximum sustainable capacity — which lets OPEC+ cut to defend a price floor against US shale growth, or unwind cuts to defend market share: a cartel of the lowest-cost barrels against a factory of the most flexible ones. No upstream thesis is complete without a view on it, because it sets the price band inside which every producer’s netback is realised.

4.6 Fiscal regimes

The last thing that decides upstream economics is the fiscal regime, the legal terms on which a government lets a company extract hydrocarbons, because the same barrel at the same price yields very different profit under different contracts (Table 23). Under a concession (royalty-and-tax) system — the norm in the United States, Canada and the United Kingdom — the company owns the oil and pays royalties and taxes; under a production-sharing contract (PSC) the state owns it, and the company recovers its costs from a slice of production (cost oil) and splits the rest (profit oil) with the state; under a service contract it is simply paid a fee per barrel.

Table 23. The three fiscal regimes

Source: fiscal-regime taxonomy from standard petroleum-economics practice and company filings; the split within each regime is negotiated per contract and varies widely.

Regimes also differ in when they charge. A US lessor royalty is taken from the first barrel (Table 9); Alberta’s Crown royalty is a flat 5% of revenue until payout, when cumulative revenue covers a drilling-and-completion cost allowance, and only then climbs on a price- and productivity-sensitive scale to 40% for oil and 36% for gas and NGLs — a burden shaped like a profit share. And because a PSC’s state share rises with profitability, the same barrel delivers less upside to a rising price than it would under a concession: reserves and production mean different things to the bottom line under different regimes, and reading a producer’s reserves without its fiscal terms is one of the sector’s classic errors.

5. Where the oil & gas is

Section 4 priced the barrel; this section maps where the barrels are. The five conditions of the petroleum system (Section 3.2) are met in only a handful of basins, and each of the great provinces satisfies them a different way.

5.1 The anchor provinces

The Middle East is where everything lined up at once: the Tethys seaway laid exceptional source rocks beside clean carbonate reservoirs under evaporite seals, over an area no other region matches, which is why supergiant fields cluster there. The Permian stacked source and reservoir rock again and again in a restricted sea, which is why an operator there drills a dozen benches beneath one lease; the Gulf of Mexico owes its traps to moving salt; and Guyana, Brazil and West Africa are one margin torn in half as South America separated from Africa — Guyana resembles West Africa because geologically it once was West Africa.

Athabasca and the Orinoco are the same accident, twice. Ordinary crude migrated hundreds of kilometres to a shallow, cool basin margin, where bacteria ate its light molecules and left bitumen behind — so the two largest “oil” accumulations on earth are conventional oil spoiled after it arrived.

Source: the petroleum-system conditions follow standard petroleum geology, as set out in Section 3.2; the Alberta oil sands’ source rocks, updip migration and biodegradation per the USGS; the rifting history and Cretaceous source rocks of the South American Atlantic margins per the USGS, 2026. Regional descriptions are the broad geological consensus rather than a basin-by-basin assessment — the basin-by-basin detail is mapped in the world oil & gas basins guide.

5.2 Production and reserves

Production and reserves tell different stories. The United States is the world’s largest producer, at a record 13.6 MMbbl/d of crude in 2025, ahead of Russia and Saudi Arabia at roughly 9.6–9.9 MMbbl/d each on the same basis (U.S. EIA). On reserves the ranking inverts, because shale reserves are produced and replaced quickly while the OPEC states’ barrels have sat in the ground for decades.

In 2010 the United States produced about 5.5 MMbbl/d and was widely assumed to be in terminal decline; horizontal drilling and fracturing more than doubled that within a decade (Figure 9). The engine is the Permian, which produced 6.6 MMbbl/d of crude in 2025 — 48% of the US total and most of its growth.

Figure 9. US crude oil production, 2010–2025

Figure data: U.S. EIA and EIA historical series; 2010 and 2015 figures approximate. The 2020 dip is the COVID demand collapse; 2024 and 2025 are successive records.

Shale rewrote the gas map too. US marketed gas reached a record 118.5 Bcf/d in 2025, two-thirds of it from Appalachia, the Permian and the Haynesville (Figure 10), and the fastest growth was Permian associated gas — gas produced alongside oil — up 11% to 27.7 Bcf/d, a major reason the region periodically runs short of pipe (Part 2).

Figure 10. US marketed natural gas by region, 2025

Figure data: U.S. EIA, 2025; “Other regions” is the residual (118.5 − 36.6 − 27.7 − 14.9). Appalachia, Permian and Haynesville together are 67% of US output.

If production is the flow, reserves are the stock, and the stock is concentrated: roughly 1,765 billion barrels of proven oil reserves at the end of 2024 on the Oil & Gas Journal’s count, about 70% of them held by OPEC members — 79% on OPEC’s own, smaller count of 1,567 billion (OPEC) — with Venezuela’s largely undeveloped extra-heavy oil nominally at the top and Canada’s oil sands lifting it into the top five (Figure 11). Gas reserves of about 7,540 trillion cubic feet on the same survey sit largely with Russia, Iran and Qatar: Iran’s alone are second only to Russia’s, and Qatar holds about 11% (U.S. EIA, U.S. EIA). The constraint has never been running out of molecules, but the price required to justify producing the next, more difficult ones.

Figure 11. World proven oil reserves by country

Source: OPEC Annual Statistical Bulletin 2025 for the six OPEC members at end-2024, with Venezuela’s, Saudi Arabia’s and Iran’s figures as reported by the Oil & Gas Journal, December 2025, which also gives Canada’s 170 billion barrels; national reporting bases differ, so the bars indicate rank. All seven except Canada are OPEC members; the group holds about 70% of world reserves on the Oil & Gas Journal’s count and 79% on OPEC’s own. Venezuela’s total is dominated by undeveloped extra-heavy oil.

5.3 Onshore vs offshore

In 2024 offshore fields produced 30% of world oil and 32% of world gas (Figure 12), and the growth inside that number is deepwater beyond 1,000 m of water, produced through FPSOs (Section 3.5), whose share of world oil output has roughly doubled since 2008 to about 6%.

Figure 12. Offshore at a glance

Figure data: Planète Énergies (TotalEnergies Foundation), 2024 shares and the deepwater series; deepwater rose from about 3% of world oil production in 2008.

The business is concentrated in a handful of provinces that behave very differently (Table 24). Brazil’s pre-salt is the largest growth engine, about 80% of Brazil’s record 3.77 MMbbl/d of crude in 2025; Guyana’s Stabroek block passed 900,000 bbl/d within six years of first oil; the US Gulf of Mexico supplies about 1.9 MMbbl/d and grows by tieback rather than new hubs; the Arabian Gulf is shallow, enormous and state-operated; and most offshore gas sits in the Qatar–Iran North Field / South Pars structure and Australia’s North West Shelf, both feeding LNG.

Table 24. The major offshore provinces

Source: US Gulf output per U.S. EIA, 2025 actual, with the share this report’s arithmetic on the 13.6 MMbbl/d US total; Guyana per ExxonMobil, November 2025; Brazil’s 2025 record and pre-salt share per Brazil’s ANP as reported by Brazil Energy Insight; global offshore shares per Planète Énergies. North Sea and West Africa figures are 2025 country totals of all liquids per the EIA’s Short-Term Energy Outlook, August 2026 — Norway 2.05 and the UK 0.77 MMbbl/d, and Nigeria’s including its onshore output — so they indicate scale rather than a like-for-like comparison.

For an investor the split is one of cash-flow shape. An offshore project takes years to first oil and commits billions before a barrel is sold, then holds a long plateau on little sustaining capital (Section 4.2), so it neither rescues a price spike nor collapses in a slump — and it concentrates value, and execution risk, in a few projects. Each offshore province’s geology and deposit parameters are mapped in the world oil & gas basins guide.

6. The companies & the value chain

6.1 The largest producers

Upstream is the least investable part of the value chain at the top, because the biggest producers are national oil companies (NOCs) — state-owned giants that are mostly closed to outside shareholders. NOCs produce about 55% of the world’s oil and gas (NRGI, 2019 estimate) and control the great majority of its reserves; the investable universe is the international oil companies (IOCs) and independents that make up the rest.

The scale gap is stark. Saudi Aramco produces around 12.9 million boe/d of total hydrocarbons (roughly 9.2 MMbbl/d of crude) — nearly three times the largest listed major, and more than ExxonMobil, Chevron and Shell together — but is over 95% state-owned. Among the investable majors, ExxonMobil leads at about 4.7 million boe/d (targeting 5.5 by 2030), followed by Chevron, Shell, TotalEnergies and BP. Below them sit the large independents — ConocoPhillips, EOG, Occidental, Canadian Natural, Diamondback — and a long tail of mid- and small-caps focused on single basins.

Table 25. Major upstream producers and their character

Sources: Aramco’s 2025 total hydrocarbon production per its Annual Report 2025, p.49, and ExxonMobil’s 2025 net production per its 2025 results, January 2026, with its 5.5 MMboe/d 2030 target per its December 2025 plan update; other production figures are approximate, company-stated and on differing bases (net vs gross, oil vs boe), so they indicate scale rather than a like-for-like ranking. Listing gives a primary exchange and ticker; NOC entries are majority state-owned and several (Russian names) are effectively inaccessible under sanctions. No market-capitalisation figures are shown by design.

Two structural features of this list matter. First, the investable majors are a minority of the industry: the reserves and low-cost production are overwhelmingly in state hands, so a public-market investor is buying the higher-cost, more marginal half of world supply. Second, the US independents are the swing producers — nimble, short-cycle and price-responsive — which is why the listed upstream universe is so heavily weighted to North American shale.

6.2 Company archetypes & structures

Upstream companies differ along two axes that decide how they behave as investments: what kind of resource they own and how much commodity-price beta that carries. Confusing a low-decline oil-sands compounder with a high-decline shale wildcatter is the most common error in the sector.

Table 26 sets out seven archetypes, from the diversified, partly integrated supermajor — the lowest-beta way in — through large-cap independents, shale pure-plays, offshore specialists, oil-sands producers and gas-weighted E&Ps, to the pre-production explorer, the highest-risk corner of the market. Each carries a different mix of commodity beta, decline and capital intensity, which Figure 13 plots.

A distinct eighth category sits apart from all of them: the mineral and royalty company, which owns the subsurface rights and collects a percentage of production revenue without paying any drilling or operating cost. Firms like Viper Energy, Texas Pacific Land, Kimbell and Canada’s Freehold and Topaz carry full commodity-price upside with none of the capital intensity — a fundamentally different risk profile.

Two quite different assets sit behind that cash-flow shape. The first is fee mineral title bought from the private owners who hold it, which puts the company exactly where the landowner stood in Table 9: it is the lessor, taking the bonus and the lessor royalty while the operator funds everything. The second is a gross overriding royalty (an ORRI in the US), carved out of somebody else’s working interest and usually bought from an operator that wants cash today. With the Crown holding 81% of Alberta’s minerals, Canadian royalty companies are built from the historic freehold plus overrides on Crown leases, while their US counterparts mostly aggregate fragmented mineral title. Both still bear production taxes, and both are wasting assets: a royalty on producing wells depletes, so acquisition pace — not production — decides whether one compounds or quietly runs down.

Table 26. Upstream archetypes and their economics

Source: archetype taxonomy from company filings and the Metal Pilot project model; examples are illustrative, not a recommendation.

Figure 13. Upstream archetypes by commodity beta and capital intensity

Source: data in Table 26; positions are the author’s qualitative assessment, not measured values, plotted one level per rating — beta: Low–medium 2, High 4, Very high and Extreme 5; capital intensity: Near zero 1, Moderate and Variable 3, High 4, Very high 5 — with tied points nudged apart inside their cell. The shaded zone marks the high-beta, high-capital quadrant where shale and offshore developers sit; mineral & royalty names are unique in carrying high beta at near-zero capital intensity.

7. Investing in upstream

Everything above converts into a way to own and read a producer: the vehicles, a reading checklist that gathers the guide’s own tests, the screen, and how the sector behaves across macro regimes, value drivers and risks.

7.1 Exposure vehicles

Unlike midstream and downstream, upstream does have a commodity an investor can buy — crude oil and natural gas trade on deep futures markets. But a futures contract is a leveraged, expiring instrument with its own well-known trap (roll cost in contango, when later-dated contracts trade above spot so every roll loses money), covered in the oil and natural gas guides; most equity investors take upstream exposure through producer shares and funds instead, and the choice among them is really a choice about how much beta and how much capital risk to carry.

The vehicle spectrum runs from most diluted to purest exposure (Table 27). Supermajor shares give the lowest-beta way in but dilute upstream with midstream and downstream; large-cap independents are cleaner E&P exposure with balance-sheet resilience; shale pure-plays and small caps are the highest-beta equities; mineral and royalty shares (Section 6.2) are the cleanest price exposure on the list — full upside, no capital treadmill, but a wasting asset that depends on somebody else’s drilling pace; and sector ETFs spread single-name risk across the group.

The depletion caveat — the upstream equivalent of contango. Every sector in this series has one structural trap, and in upstream it is paying for barrels that deplete. A producer’s earnings look cheapest at the top of the cycle, exactly when the value of every future barrel is about to fall, and high current output can mask a fast-declining base that will cost a fortune to replace — so value a producer on mid-cycle prices and full-cycle F&D cost, not on trailing earnings at a peak price. Investors who bought high-decline producers on peak-2022 cash flow and watched both the price and the production fall learned this the expensive way.

Table 27. Ways to take upstream exposure

Source: vehicle structures from issuer documentation and company filings; structural attributes only — no fees, yields or prices are quoted. Mineral & royalty consolidation is active — Viper Energy’s ~$4.1 bn purchase of Sitio, including Sitio’s net debt (Viper Energy, June 2025), is the largest recent example.

7.2 How to read & screen a producer

An upstream balance sheet looks nothing like a pipeline’s or a refinery’s. The defining asset is reserves, and the defining expense is the depletion of them — the accounting recognition that every barrel sold permanently shrinks the asset — so where midstream is screened on contracts and downstream on complexity, upstream is screened on reserves, cost and balance sheet: those decide whether a producer survives the next downturn and compounds through the next upturn. The guide’s own tests gather into one reading order (Table 28) — what the company produces and owns first, then what it costs, then how fast it runs down, and only then the balance sheet and the valuation. It is a framework for reading a producer, not advice, and the vehicle chosen in Section 7.1 decides how much of it applies.

Table 28. The upstream reading checklist — where each check comes from

Source: this guide’s own checks, developed in the sections cited, which carry the underlying sources; a reading framework and an ordering of steps, not a measured ranking or advice.

Seven measures do most of the work, and they are all in the filings — Section 4.1 shows where each cost line sits and how to put two companies on one basis.

Reserve life (the reserve-life index, RLI: proved reserves divided by annual production) and the PDP share come first — deep, largely-producing reserves are durable; a short reserve life is a warning. Breakeven and full-cycle cost decide survival: a producer that profits at $45 oil lives through a bust that kills one needing $70. F&D cost and reserve-replacement ratio measure whether the company creates value as it drills or merely runs to stand still. Netback per boe captures the realised profit at the wellhead after transport and royalties. Net debt to EBITDA matters enormously because leverage on a volatile revenue line is what turns a downturn into a bankruptcy — the producers that entered 2020 with low debt are the ones that survived it. Free-cash-flow breakeven — the price at which the company funds both its capital programme and its dividend from operations — is the modern discipline test. And the recycle ratio (netback divided by F&D cost) is the single cleanest measure of whether a producer’s drilling actually creates value.

The valuation anchor is the standardised measure and PV-10 — the present value of future net cash flows from proved reserves, discounted at 10% under a prescribed price — the SEC’s attempt to put a single dollar value on the asset, and highly sensitive to the price assumption baked into it. Valuation itself is done on cash-flow multiples (EV/EBITDAX — enterprise value to cash earnings before exploration expense), on EV per barrel of 2P reserves, and on P/NAV — price to a discounted, PV-10-style net asset value. Those ratios are the screening machinery rather than facts to print in an evergreen guide, which is why this report names them without quoting levels — you can screen the upstream universe on Metal Pilot and, because a producer’s realised price depends on getting its barrels to market, cross-check the pipelines and terminals that evacuate them at the same time.

Table 29. Screening metrics for upstream companies

Source: metric taxonomy from upstream company filings and reserve reports; the Metal Pilot project model captures the asset-side inputs — production, reserves, resource base — behind several of these.

7.3 Macro regimes, rates & correlations

Upstream’s macro profile is the simplest in the chain: it is the levered call on the commodity that Section 1.5 described, and everything else is second order. Rates matter through the cost of capital for capital-intensive developers and the discounting of long-dated reserves, but upstream is far less bond-like than midstream. Cost inflation matters more, because a producer’s real exposure is the spread between price and cost: the 2021–23 surge in oilfield-service, steel and labour costs lifted breakevens across the industry and quietly eroded the shale cost advantage.

The worked episode — upstream amplifies, in both directions. The cleanest concluded test spans 2020 to 2022. In 2020, the pandemic demand collapse drove the expiring May WTI futures contract to an unprecedented settlement of −$37.63/bbl on 20 April 2020, and a wave of US producers — Chesapeake, Whiting, Denbury among them — filed for bankruptcy protection as their leveraged, high-decline models met a price shock. Barely two years later, the 2022 price spike after Russia’s invasion of Ukraine handed the survivors the best cash flows in their history; Occidental returned roughly +119% on the year and was the best-performing stock in the S&P 500, and the broad E&P group ran far harder than the barrel itself. That is the signature of the sector: it does not merely track the oil price, it levers it — spectacularly on the way up and brutally on the way down.

Table 30. Upstream across macro regimes

Source: regime responses substantiated by the episodes cited; the May 2020 NYMEX WTI futures settlement per the CFTC, November 2020; equity moves are widely reported for the windows shown. Historical, not predictive; correlations are regime-dependent. Cross-commodity regime behaviour is covered in the macro-regime guide.

7.4 Value drivers & cycles

Several forces move upstream economics (Table 31). The commodity price sets the ceiling and the floor and dominates all the others, and decline sets how hard a company must run to stand still (Sections 4.3–4.4). The driver that has changed most is capital discipline: after a decade of debt-funded, growth-at-any-cost drilling that destroyed capital, the US shale industry pivoted around 2020 to “shale 2.0” — living within cash flow, returning cash to shareholders through dividends and buybacks, and growing production only modestly — which is the single biggest reason the sector took the 2020 shock and the 2022 windfall so differently from how it would have a decade earlier. Drilling productivity — longer laterals, better completions, more wells per rig — keeps lowering breakevens even as the easiest rock is used up, which is why US production kept rising in 2025 on 5% fewer rigs.

The consolidation wave is the clearest recent illustration of the sector maturing. Beginning in late 2023, roughly US$300 bn of US upstream M&A — a record US$192 bn in 2023 and US$105 bn in 2024 (Enverus) — remade the industry, with US$134 bn worldwide in 2024 alone (Wood Mackenzie), as majors and large independents bought scale and inventory rather than drilling for it.

Figure 14. The upstream consolidation wave — largest deals

Source: the announcements of ExxonMobil–Pioneer (October 2023), Chevron–Hess (October 2023), Diamondback–Endeavor (February 2024), ConocoPhillips–Marathon (May 2024) and Occidental–CrownRock (December 2023); every value is the announced transaction value including the target’s net or assumed debt — for Chevron–Hess the US$60 bn total including debt, against US$53 bn for the shares alone. Total disclosed global upstream M&A was US$134 bn in 2024 across 189 deals, down 42% on 2023, per Wood Mackenzie.

Table 31. Upstream value drivers

Source: drivers derived from the concluded episodes cited in Sections 4 and 7.3 and from company capital-allocation disclosures, 2014–2026.

7.5 Risks, controversies & ESG

Upstream’s risks are the most numerous in the value chain, and Table 32 registers them. The biggest is the one the whole business is built on: a price-taker with operating leverage is exposed to every fall in oil or gas, and that risk has bankrupted more E&P companies than all the others combined — the 2020 episode in Section 7.3. Reserves can be revised down, service costs inflate, every producing asset carries a growing asset-retirement obligation, and a government can change the fiscal terms or impose a windfall tax (Section 4.6). Two deserve more than a row.

Water and induced seismicity bite commercially. Every barrel of oil brings up several barrels of salty produced water — Texas alone generated roughly 3.9 billion barrels of it in a year — and injecting it into deep disposal wells has been linked to earthquakes: after a magnitude 5.4 event in Reeves County in November 2022, the Texas Railroad Commission cut deep disposal in the area from 298,000 to 162,000 barrels a day, and in early 2024 it suspended the permits of 23 disposal wells. Losing disposal capacity raises the cost of every barrel and can cap production in an area however many locations a company holds.

Methane and peak demand are the long-horizon controversies, and both are two-sided. Methane leakage from wells and gathering is upstream’s sharpest near-term ESG exposure, increasingly regulated and measured. The structural question is peak demand: the bear case is that long-life reserves face a demand plateau inside their economic life, the bull case that developing-world, aviation and petrochemical demand runs for decades while under-investment keeps prices — and the survivors’ cash flows — high. Both rest on assumptions about 2040s demand that no one can verify, and the IEA and OPEC disagree by tens of millions of barrels a day (Section 8).

Table 32. Upstream risk register

Source: risks derived from the concluded episodes cited above — the 2014–16 and 2020 busts and bankruptcies, the 2022 service-cost inflation, and the peak-demand debate in Section 8. Likelihood and impact are the author’s qualitative assessment, not a modelled probability.

Figure 15. Upstream risk map

Source: data in Table 32. Each point prints its likelihood × impact score, with one level per rating — likelihood High 4 and Medium 3, with Uncertain plotted mid-scale at 3; impact High 5 and Medium 3 — and tied points nudged apart inside their cell; the shaded region is the high-likelihood, high-impact quadrant. Ratings are the author’s assessment, not disclosed figures.

8. Future outlook & catalysts

Three forces will decide upstream’s next decade, and for the first time they pull in genuinely different directions.

Consolidation and capital discipline are reshaping the investable universe. The mega-merger wave (Section 7.4) has concentrated the best US inventory in a handful of majors and large independents, and the “shale 2.0” ethos of returning cash rather than chasing growth has held even through a strong price environment. The result is a more mature, more profitable, slower-growing US sector — one that behaves more like a cash-return machine and less like a growth story than it did a decade ago. Whether that discipline survives the next price spike is the open question.

The demand outlook is the defining and most contested backdrop. The IEA expects oil demand to grow only modestly — about 2.5 MMbbl/d between 2024 and 2030 to a plateau near 105.5 MMbbl/d, with China peaking around 2027 and EVs displacing more than 5 MMbbl/d by decade’s end. OPEC sees the opposite: demand climbing to 113 MMbbl/d by 2030 and past 120 MMbbl/d by 2050, driven by developing-world growth, aviation and petrochemicals. That gap — tens of millions of barrels a day — is the single largest uncertainty hanging over every long-life upstream investment, and it explains why low-cost, short-payback barrels are increasingly favoured over decade-long megaprojects.

New demand is pulling gas in directions that did not exist last cycle. US gas is being drawn to LNG export (feeding the midstream build-out in Part 2) and, for the first time, to data-centre power as electricity demand grows for the first time in two decades. Gas-weighted producers in Appalachia and the Haynesville are the most direct upstream beneficiaries. Meanwhile the offshore renaissance — Guyana’s Stabroek block above all, plus Brazil’s pre-salt and a revived Gulf of Mexico — is delivering some of the lowest-cost, longest-life new barrels in the world, a reminder that the conventional half of the industry is far from finished.

Table 33. What to watch

Source: demand outlooks per IEA Oil 2025 and OPEC’s World Oil Outlook 2025 (almost 123 mb/d by 2050), with its 113.3 mb/d for 2030 as reported by Argus, July 2025; US production and productivity per U.S. EIA. Timings are agency projections and slip routinely.

9. Summary

Upstream oil and gas is the finding-and-producing end of the hydrocarbon chain — exploration, appraisal, development and production — and it earns by selling the raw commodity at the wellhead, not a fee or a margin. Because revenue is simply volume times price minus the cost of lifting and finding the barrel, upstream amplifies the commodity where midstream dampens it. What it owns in the ground is a graded claim — proved, probable or possible, booked or merely inventory — and the gap between deep producing reserves and a large unbooked story is where most upstream mistakes are made.

Behind every cost figure is a physical path — a lease that must be drilled before its clock runs out, a shale well that costs about $7–8m and a deepwater hub that costs up to $10bn, a meter where title passes — and its shape, whether a short-cycle shale factory, a sunk deepwater hub or a near-zero-decline oil-sands plant that can upgrade the heaviest barrel into a light one, decides how a producer behaves when the price moves. The survival price is a stack: operating cost, the royalty and midstream claims ahead of the netback, and the sustaining capital that decline demands. A new Permian well starts near 1,100 boe/d (derived) and, on a modelled curve, gives up roughly a quarter of its lifetime barrels in its first year, so reserve replacement and F&D cost decide whether a producer compounds or quietly liquidates.

Shortage shows up as a rising price and a rising breakeven, not a spread, and the coordinating force is OPEC+, not a regulator. The barrels sit mostly with OPEC states even as the United States produces the most, more than half the world’s output comes from national oil companies closed to investors, and the listed remainder sorts into archetypes from the low-beta supermajor to the pure-beta wildcatter. The forward case turns on capital discipline, consolidation and the peak-demand debate; the disciplined way to own it is on mid-cycle prices and full-cycle cost, read in the order of the checklist in Section 7.2 — because in upstream, the same price move that makes a producer soar is the one that can wipe it out.

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

Next in the series → Part 2: Midstream — how the barrel is moved, processed and stored once it leaves the wellhead.

10. Vocabulary

Table 34. Upstream terminology

Source: definitions compiled from U.S. EIA, SPE-PRMS and U.S. SEC reserves rules, company filings and standard industry usage.

11. Sources, methodology & disclaimer

11.1 Sources, methodology & data vintage

Agencies and regulators. U.S. EIA — the United States produced more crude oil than any other country in 2025 and the 2025 crude production record; U.S. EIA — US natural gas production reached a new record in 2025; U.S. EIA Short-Term Energy Outlook, August 2026 for world and country liquids output in 2025; the CFTC’s interim staff report on NYMEX WTI trading around 20 April 2020; SEC modernized oil and gas reserves reporting rules; SEC Regulation S-X Rule 4-10(a) and Regulation S-K Item 1202 for the reserves definitions and disclosure requirements; NI 51-101 and CSA Staff Notice 51-324 for Canadian reserves reporting; Dallas Fed Energy Survey, Q1 2026 (breakevens); OPEC Annual Statistical Bulletin 2025 and its launch release, and the World Oil Outlook 2025 launch release; U.S. EIA country analysis of Iran and of Qatar’s gas for gas-reserve rank and share.

Capital costs. Permian Resources Q3 2025 and Q4 2025 results for shale drilling and completion cost per lateral foot; Transocean fleet status report, July 2025, for drillship day rates; World Oil for the Whiptail subsea contract values; Offshore Technology and ExxonMobil for Yellowtail’s sanction value and start-up; International Petroleum Corporation for Blackrod Phase 1.

Exploration and by-products. Westwood Energy — The State of Exploration 2026 (high-impact commercial success rates, ultra-deepwater results and well counts); USGS Mineral Commodity Summaries 2026 — sulfur and 2025 — helium for recovered by-product volumes; The American Oil & Gas Reporter for Permian water cuts and produced-water volumes.

Drilling, completion and offshore. Offshore structure types and their deepest installations as of 2005 per NOAA Ocean Explorer; the Turritella FPSO per SBM Offshore, 2016; subsea tieback economics per Offshore Engineer, 2024. Water volumes per fractured well per the USGS; fracturing-fluid composition and additive functions per FracFocus; lateral-length shares and the flat proppant and fluid intensity since 2017 per U.S. EIA, 2021; the West Texas seismic response area and disposal cuts per the Railroad Commission of Texas, December 2022, with the 2024 permit suspensions and Texas produced-water volumes as reported by the Texas Tribune, January 2024. U.S. EIA — longer wells boost Permian crude oil and natural gas production (lateral lengths and super-laterals, 2025); U.S. EIA — natural gas delivery and storage (separation, gathering, dehydration, amine treating, nitrogen rejection, cryogenic separation and fractionation); U.S. EIA — Gulf of America production; ExxonMobil Guyana (Stabroek at 900,000 bbl/d, November 2025); Brazil Energy Insight reporting Brazil’s ANP 2025 production record and pre-salt share; Planète Énergies (TotalEnergies Foundation) for the offshore and deepwater shares of world production and reserves; ExxonMobil and Hess (Payara scope and start-up); BSEE (the Well Control Rule).

Land tenure and leasing. BLM — onshore oil and gas leasing rule fact sheet (bids, rentals, royalty) and the Mineral Leasing Act, 30 U.S.C. § 226 (the 10-year federal primary term) and BLM — Interior implements the One Big Beautiful Bill Act (the 12.5% royalty floor); BOEM Gulf of America lease sales; New Mexico State Land Office on the 25% state royalty; Texas Mineral Interest Pooling Act (Natural Resources Code ch. 102); Alberta.ca — about tenure and tenure overview; CAPP — mineral rights; C-NLOER land-rights issuance process; and Alberta’s Modernized Royalty Framework overview (the 5% pre-payout rate and the post-payout scales).

Geology and provenance. The Alberta oil sands’ source rocks, migration and biodegradation per the USGS; the rifted-margin history and Cretaceous source rocks of the South American margins per the USGS, 2026.

Industry bodies and market data. SPE Petroleum Resources Management System (PRMS) for reserves classification, with the SPE PRMS FAQs on its five-year benchmark, forecast pricing and resource sub-classes; the Oil & Gas Journal worldwide reserves survey, December 2024, for world oil and gas reserves and OPEC’s share on that count, with its December 2025 survey for the Canadian and leading OPEC country figures; Rystad Energy for average new-project breakevens by supply segment; IEA Oil 2025 for the demand outlook; Natural Resource Governance Institute for NOC scale; raw bitumen’s sulphur per Oil Sands Magazine, SAGD recovery per Oil Sands Magazine and the mining plant-recovery requirement per the Alberta Energy Regulator (Directive 082); recovery factors and the US enhanced-oil-recovery mix per the U.S. Department of Energy with shale recovery efficiencies per the U.S. EIA; oil-sands upgrader yields per the Canada Energy Regulator and the hydrocarbon gas liquid components and ethane rejection per U.S. EIA; crude-quality classification per the Wyoming State Geological Survey, the WTI deliverable specification per CME Group, Canadian heavy-crude assays per CrudeMonitor and bitumen and dilbit specifications per Oil Sands Magazine; S&P Global Commodity Insights (June 2025) and Enverus Intelligence Research (April 2025) for oil sands half-cycle and greenfield breakevens.

Operating costs and cost disclosure. Per-unit 2025 costs per Canadian Natural (Q4 2025 report), Cenovus Energy (2025 annual MD&A), Suncor Energy (Annual Report 2025), Diamondback Energy (FY2025 results), EQT (Q4 2025 results), Saudi Aramco (Annual Report 2025) and Petrobras (Performance Report 4Q25); produced-water pricing per Aris Water Solutions (FY2024); sour-gas treating costs per SPE PetroWiki; disclosure rules per Regulation S-K Item 1204, Regulation S-X Rule 4-10, FASB Statement No. 69, Form 51-101F1 and the OSC 2014 amendments to NI 51-101.

Decline rates. Field-by-field observed and natural decline rates, the no-investment scenario and the shale output profile per the IEA, 2025; US onshore production lost from existing wells per the U.S. EIA, 2025; Permian first-year well declines per JPT, 2019; the deepwater profile per BOEM; coalbed methane per SPE PetroWiki; terminal decline per the Payne Institute; oil-sands profiles per the Canada Energy Regulator; corporate decline rates per Canadian Natural and Peyto.

Per-well rates and recoveries (Table 22a). US well counts, the stripper share and the 15 boe/d threshold per U.S. EIA and U.S. EIA; Permian new-well productivity per U.S. EIA, 2022, and first-full-month volumes per U.S. EIA, 2024; play-average EURs for unproved undrilled locations per U.S. EIA (AEO2023) and the 30-year well life per U.S. EIA (AEO2026); shale-gas well rates per Comstock Resources and Antero Resources, 2026, with Rystad Energy’s per-foot first-year median reported by OilPrice, 2025; Guyana well counts and capacities per ExxonMobil; Brazilian pre-salt volumes and well counts per ANP, June 2026; deepwater EUR per Wood Mackenzie, 2024; Saudi per-well rates per JPT, 2015, and the ANDR-1 well per Aramco, 2025; SAGD well-pair counts and production per Suncor Energy, 2024 AIF, Oil Sands Magazine and Natural Resources Canada, 2024.

Company sources. Saudi Aramco, ExxonMobil, Chevron, Shell, TotalEnergies, BP, ConocoPhillips, EOG, Occidental, Canadian Natural, Diamondback and Petrobras disclosures and websites, 2025–26, for production, reserves and strategy; ExxonMobil’s 2025 results and 2030 plan update; deal values from the ExxonMobil–Pioneer, Chevron–Hess, Diamondback–Endeavor, ConocoPhillips–Marathon, Occidental–CrownRock and Viper–Sitio announcements, and US upstream deal totals per Enverus; 2025 production volumes, product mix and realised prices per EQT and Canadian Natural, Q4 2025 reports, the latter also the source of the standard Canadian BOE advisory; ExxonMobil’s oil-sands reserve de-bookings and rebooking per IEEFA and IEEFA; EOG’s premium and double-premium well definitions per Hart Energy.

Methodology. Production and reserve figures are taken from agency and company reports at the latest published year for each series; because those publish on different cycles, several figures carry different vintages and are labelled where material. Country production figures use differing bases (crude vs total liquids) and country reserve figures follow national reporting, so both indicate rank rather than a like-for-like ranking; Venezuela’s reserves are dominated by undeveloped extra-heavy oil. Company production is company-stated on mixed net/gross and oil/boe bases. Breakevens are survey averages that move with cost and efficiency and should be re-sourced. Per-barrel operating costs in Section 4.1 are company-reported for 2025 on each company’s own definition and in its reporting currency, so they order the types of barrel rather than rank them like-for-like; the produced-water cost per barrel of oil is this report’s arithmetic on one company’s handling price and the basin’s water ratio. Decline rates in Section 4.3 mix measured and modelled figures — the IEA’s observed post-peak rates are measured, its natural-decline and shut-in figures and the BOEM and CER profiles are modelled — and the US base decline of close to 40% is this report’s arithmetic on the EIA’s existing-well loss against December 2024 output. Per-well rates and recoveries in Table 22a sit on mixed bases — first-month, first-three-month, 60-day and annual averages, oil-only and boe, design capacity and measured output — so they order the types of well rather than benchmark them, and every cell marked derived is this report’s arithmetic on the two inputs named beside it. The shale EUR range is modelled on an Arps curve scaled to the EIA’s new-well rate and run to the 30-year life the EIA assumes; the agency play-averages beside it cover a play’s remaining undrilled locations rather than the wells an operator drills first, and the two are not comparable. The per-flowing-barrel figures in Section 4.3 divide the well cost of Table 12 by the rate on the line. The cost-curve ranges in Table 21 are indicative orderings, not precise figures. Leasing terms in Section 3.1 are statutory or regulatory as of the data date — the US federal royalty floor moved twice between 2022 and 2025 and the BLM has further fee and rental revisions in progress — so treat every rate, bid minimum and rental there as a dated starting point to re-source; private-lease bonuses and royalties are negotiated and unpublished, and the ranges given are market convention rather than measured averages. Demand outlooks — the IEA plateau and the OPEC growth path — are agency projections and are labelled as such wherever they appear. No market capitalisations or share prices appear anywhere in this report, by design; the deal values in Section 7.4 are announced transaction values.

Data as of 26 August 2026. Intended update cadence: annually, after the EIA production records and the IEA/OPEC demand outlooks.

11.2 Disclaimer & disclosure

This report is for informational purposes only and is not investment advice, a recommendation, or an offer to buy or sell any security. Historical price and equity behaviour is presented as settled history; past performance is not indicative of future results, and the correlations and regime patterns described here are sample-dependent and can break down, particularly in a crisis. Readers should do their own research and consider consulting a licensed financial adviser before acting. This report was prepared with the assistance of AI, consistent with the ai = true flag in its front matter; figures were sourced from the references above and reviewed, but readers should verify any figure they intend to rely on. The author holds no position in any company named in this report.