In a companion note, we examine how contract typology determines V&V exposure across EPC and EPCM engagements: Why 70–90% of EPC/EPCM Projects Fail — and What Contract Structure Actually Controls. This note extends that argument to the capital dimension.
The number everyone trusts
Owners rank capital projects by breakeven. It is the wrong first question. Two developments with the same breakeven can sit at opposite ends of defensibility, because breakeven says nothing about how the capital was structured to deliver — or whether the owner can recover value when delivery fails. The variable that carries that information is execution structure: whether a development is concentrated into a single managed interface or atomised across many. Capital intensity is where that structure first becomes visible — and, as a companion to our note on V&V exposure across contract typologies, it is where the typology decision starts to show its cost.
Measuring concentration
Capital intensity — CAPEX divided by installed capacity, in dollars per barrel of oil per day — is a blunt instrument, but a revealing one. Consider one basin, one operator, seven developments.
| Development (FPSO) | FID | CAPEX (gross) | Capacity | CAPEX/bopd |
|---|---|---|---|---|
| Liza 1 (Destiny) | 2017 | $4.4B | 120k | $36.7k |
| Liza 2 (Unity) | 2019 | $6.0B | 220k | $27.3k |
| Payara (Prosperity) | 2020 | $9.0B | 220k | $40.9k |
| Yellowtail (ONE GUYANA) | 2022 | $10.0B | 250k | $40.0k |
| Uaru (Errea Wittu) | 2023 | $12.7B | 250k | $50.8k |
| Whiptail | 2024 | $12.7B | 250k | $50.8k |
| Hammerhead | 2025 | $6.8B | 150k | $45.3k |
| Programme (7) | ~$61.6B | ~1.46 Mbopd | ~$42k |
Normalisation note: The $12.7B figure for Uaru and Whiptail reflects a standardised cost model for the 250k bopd FPSO platform; identical pre-FID estimates are consistent with ExxonMobil's disclosed programme economics for that capacity class. Liza 1 and Liza 2 include FPSO lease capitalisation that later headline figures may not; Liza 2 and Payara have since been debottlenecked to approximately 265k bopd each, reducing their effective cost per barrel below nameplate. Nameplate capacity is used throughout for cross-asset comparability. Pick one basis and hold it.[1] [2]
Each development is a single FPSO — a near-turnkey package with one dominant interface. The programme commits roughly $61.6 billion for about 1.46 million barrels per day: some $42,000 per flowing barrel, and, per development, one interface to govern.[3] [4] [5] [6]
Why capital intensity is not destiny
Read across segments and the intuition breaks.
| Segment / asset | Type | CAPEX/bopd | Note |
|---|---|---|---|
| Brazil — Búzios | Greenfield deepwater | ~$18–22k | ~$4B/FPSO (early units), 180–225k bopd |
| GoM — Shenandoah | Greenfield deepwater | ~$18k | $1.8B / 100k bopd |
| Guyana — Stabroek (programme) | Greenfield deepwater | ~$42k | one-time per FPSO |
| GoM — Anchor (20k psi HP/HT) | Greenfield deepwater HP/HT | ~$76k | $5.7B / 75k bopd |
| Venezuela — Orinoco | Brownfield / restoration (projected) | ~$33–60k | risk- & overhead-inflated |
| Kazakhstan — Kashagan | Greenfield, multi-party | ~$145k | $55B / 380k bopd, Phase 1 |
| Canada — oil sands | Greenfield | ~$137k (est.) | contrast |
Capital intensity does not track greenfield versus brownfield, or offshore versus onshore.[9] [11] [12] [13] [14] A prolific greenfield (Búzios) is cheaper per barrel than Guyana; an onshore play (the Permian) is a recurring spend rather than a one-time one — roughly $12,000 per flowing barrel every year to hold output flat, against a few thousand wells drilled annually out of tens of thousands cumulatively; and a greenfield (Kashagan) is the most capital-intensive of all. Complexity comes from two places, and only one is a decision. The first is exogenous — the stakeholders and geology a project inherits: communities, land, social licence, reservoir behaviour. Offshore carries far less of it; a deepwater FPSO has no surface community, and the counterparties reduce to the state, the partners and the regulator. Onshore, in weak-institution jurisdictions, disperses accountability across actors no contract can concentrate. The second is endogenous — how the owner splits the execution: one integrated interface or many. This one is chosen, and offshore does not choose it for you — Kashagan and Cantarell are both offshore, and both atomised. The allocator inherits the first and decides the second; concentration and reservoir quality pull intensity down, while multiplied interfaces, dispersed stakeholders and decline push it up. Most of the contest is decided in the second.
Productive capital and destructive capital
Because capital intensity is not destiny, the objective is not to minimise it. It is to distinguish productive capital from destructive capital. The two can sit on an identical CAPEX line and behave in opposite ways.
Some capital intensity is bought on purpose. Anchor's roughly $76,000 per barrel is the price of 20,000-psi high-pressure, high-temperature engineering — the materials, redundancy and containment a hostile reservoir requires. The barrel is expensive because the physics is, and the spend buys reliability, maintainability and safety by design. That is productive capital: cost that returns as lower operating risk and life-cycle cost even as it raises the headline CAPEX. A CFO is right to say that more rigorous engineering does not always reduce CAPEX — sometimes it deliberately raises it. That objection is correct, and it describes productive capital, not waste.
Other capital intensity is the residue of decisions. Kashagan's roughly $145,000 was not bought; it accumulated — through late change, undefined interfaces, rework and accountability dispersed across seven parties. That is destructive capital: spend that buys claims, delays and change orders instead of capacity. Two developments can carry the same CAPEX and the same breakeven while one is mostly productive and the other mostly destructive. The headline number does not tell you which. The execution structure does. The distinction that governs the outcome is not high versus low capital intensity — it is productive versus destructive.
Two ceilings: Kashagan and Cantarell
Two assets mark the limits, and neither is a reservoir story.
Kashagan is the limiting case of atomised governance in a hostile environment: roughly $55 billion for 380,000 barrels per day in its first phase — about $145,000 per flowing barrel — developed by a seven-company consortium in sour, high-pressure crude that cracked the export pipelines.[8] The overrun, which pushed lifetime cost estimates well past a hundred billion dollars, was not a geology problem. It was an interface problem, at national scale — precisely the failure mode the N2 and V&V-exposure frameworks anticipate when accountability is distributed across many parties before a single deliverable is assessed.
Cantarell is the limiting case of atomised execution on a mature asset. To arrest its decline, Pemex ran the world's largest nitrogen-injection project across a contractor base numbering in the hundreds — over two hundred firms in the complex, more than ninety per cent foreign, with external project management brought in to coordinate them. The capital did not fail on geology; it failed on an interface. The injection system was scaled without a matching investment in gas handling: nitrogen migrated into the produced gas, collapsed its value, and forced flaring the regulator later valued in the billions — tied to more than a hundred thousand barrels a day left in the ground. Pemex spent close to $2.9 billion in a single year on new wells and processing and committed some $6 billion to hold production near 325,000 barrels per day; output still fell to roughly 159,000 by the end of the decade, down from a peak above two million.[10] A depleting reservoir raises the marginal barrel's cost; an unmanaged interface between capital systems multiplies it — and almost none of it shows in operating metrics until the decline is irreversible.
Two breakevens
This is where country and project risk enter, and where a common error begins. Capital intensity feeds the cost breakeven directly. Country and project risk do not move the cost breakeven; they move the hurdle. Venezuela's Orinoco is cheap to lift but expensive to finance: restoration overhead and risk push the sanction breakeven above $80 per barrel, and a majority of the required capital needs sustained $80 oil to clear.[7] Risk converts a cheap barrel into an unbankable one. Any forensic assessment must separate the cost breakeven from the sanction, or hurdle, breakeven — they are set by different things and fail for different reasons.
Where the capital hides
None of this is visible in real time. High capital intensity surfaces in the accounts: as depreciation, depletion and amortisation per barrel — spending spread across reserves or units, quietly compressing margin — and, when carrying value exceeds the recoverable amount, as impairment. The value destroyed by an over-capitalised, atomised, badly-typed engagement does not announce itself operationally. It appears in DD&A and in write-downs, after the decision can be changed. Destructive capital, in other words, is recognised late and by the accountants — which is exactly why the analysis belongs at contract award, not at year-end.
The typology owners don't buy
Which returns the argument to typology. Environmental and institutional conditions — local-content obligations, social licence, reservoir decline, hostile geology — push owners away from a single concentrated, near-turnkey interface and toward atomised, multi-contractor structures. That is often the right call for the operation. But the same atomisation applied to implementation multiplies interfaces, raises capital intensity and disperses accountability — and owners then try to extract a concentrated outcome from a dispersed instrument: demanding improvement in production indicators under a staff-augmentation contract that carries no results obligation and was never priced for one. The most value-destroying position in any portfolio is the high-intensity, atomised barrel governed by a typology the owner did not buy. The question at contract award is not only "what will this cost?" It is: does the typology match the execution structure our environment forces on us — and are we paying for the obligation we intend to demand?
Verification and validation as a capital discipline
Seen this way, verification and validation is not a quality-assurance step bolted on at the end. It is a capital-allocation discipline. Every ambiguous requirement, every undefined interface, every late change is a financial decision before it is a technical one. V&V does not guarantee a good outcome — nothing does — but it increases the probability that capital is allocated where it creates value rather than where it absorbs uncertainty. Productive investment emerges when requirements, interfaces and execution structure reinforce one another; destructive investment emerges when uncertainty is transferred between them rather than resolved — and V&V is the discipline that keeps that transfer from going unpriced.
The deeper answer is not to retreat toward the simplest barrel. The most complex developments are among the most profitable when they are governed — Guyana is complex and world-class at once. The differentiator is not avoidance; it is capability: disciplined requirements and interface management — the N2 and V&V-structural practice that declares every interface before the first contract is awarded. The industry has been slow to make this a default, least of all in our region. Avoiding it is declining the tropics for want of a vaccine — leaving the most rewarding ground to whoever bothered to equip for it.
None of this begins at the construction site. The ability to influence a project's cost is greatest at the front end and collapses as the work proceeds — the relationship Patrick MacLeamy's curve made familiar in design, and that Edward Merrow's analysis of hundreds of industrial megaprojects made quantitative in the capital-project record: front-end loading at the final investment decision is the single most reliable predictor of outcome, and where it is weak, cost deviations of fifty per cent or more follow.[15] [16] Capital is rarely destroyed during construction. It is usually destroyed much earlier — in requirements, architecture, interfaces and execution strategy, before the first contract is even awarded. Engineering determines whether that spending becomes productive investment or irreversible waste.
JR Engineering Company applies SE-grade V&V and interface management to capital-project structuring across oil & gas and energy engagements — verification that every contracted obligation and every system interface is declared in a form that can be tested, before the capital is committed rather than after the impairment.
References
- OGJ / NS Energy. ExxonMobil makes FID on Liza development offshore Guyana; Liza Phase 1 overview. ogj.com; nsenergybusiness.com. (Liza 1: $4.4B / 120k bopd.)
- Hess Corporation. Hess Sanctions Liza Phase 2 Development Offshore Guyana. investors.hess.com. (Liza 2: $6.0B / 220k bopd.)
- ExxonMobil. ExxonMobil to Proceed with Payara Development Offshore Guyana. corporate.exxonmobil.com, 30 September 2020. (Payara: $9.0B / 220k bopd.)
- JPT / SPE. ExxonMobil Brings Fourth FPSO Online Offshore Guyana. jpt.spe.org. (Yellowtail ONE GUYANA: $10.0B / 250k bopd.)
- ExxonMobil. ExxonMobil Guyana Expands Capacity. corporate.exxonmobil.com, 22 September 2025. (Hammerhead: $6.8B / 150k bopd.)
- Kaieteur News Online. Guyana's breakeven and programme economics. kaieteurnewsonline.com, 13 January 2025. (Breakeven ~$28–36/bbl; productive wells <$20.)
- Oil Price / AJOT / Dermody, R. Trump's Venezuela Oil Plan Runs Into Hard Reality; What Would It Take to Bring Venezuela's Output Back to 3 Mbopd; The Heavy Crude Realignment. oilprice.com; ajot.com; rdermody.substack.com. (Orinoco: ~$33–60k/bopd; hurdle breakeven >$80; Canada oil sands ~$137k/bopd.)
- NS Energy / OilPrice / Eurasian Research. Kashagan Oil Field Development; World's Most Expensive Oil Project; Kashagan field study. nsenergybusiness.com; oilprice.com; eurasian-research.org. (Kashagan Phase 1: ~$55B / 380k bopd ≈ $145k/bopd; lifetime estimates $116–136B.)
- Discovery Alert / Ecoticias. Brazil's Pre-Salt Deepwater Revolution 2025; Búzios field overview. discoveryalert.com.au; ecoticias.com. (Búzios: ~$4B/FPSO, 180–225k bopd → ~$18–22k/bopd.)
- Offshore Technology / Wikipedia / Romo, D. (El Colegio de México, The Cantarell Oil Field and the Mexican Economy) / Offshore Magazine / CNH. Cantarell Oil Field Development; Cantarell Field; Pemex to invest $6B in Cantarell production. offshore-technology.com; en.wikipedia.org; offshore-mag.com. (Cantarell: ~$2.88B in 2008 on wells and processing plants; ~$6B committed to sustain output; 200+ contractors, >90% foreign, with external project management; nitrogen–gas-handling interface failure → ~$1.3B flaring losses 2009–12 and ~106k bopd unextracted per CNH; output fell to ~159k bopd by 2019 from a peak above 2 Mbopd in 2003–04.)
- JPT / SPE. Gulf of Mexico Operators Transition to Faster, Cheaper Developments. jpt.spe.org. (Shenandoah/Beacon: $1.8B / 100k bopd → ~$18k/bopd.)
- Chevron / JPT. Chevron Sanctions Anchor Project; Chevron's $5.7 Billion Anchor Project Starts Production. chevron.com; jpt.spe.org. (Anchor 20k psi HP/HT: $5.7B / 75k bopd → ~$76k/bopd.)
- Permian Resources / EnergyNewsBeat / Peak Oil Barrel. Strong Fourth Quarter 2024 Results; Survey note; Permian Basin shale gas. permianres.com; energynewsbeat.co; peakoilbarrel.com. (Permian sustaining: ~$1.9–2.1B/yr for ~170k bopd; ~$12k/bopd/yr.)
- WhatIsPiping / Novap Services. Brownfield vs Greenfield Development. whatispiping.com; novapservices.com. (Conceptual distinction: brownfield reuses infrastructure; marginal brownfield barrel can be the costliest.)
- MacLeamy, P. / Construction Users Roundtable (CURT). The MacLeamy Curve — design effort versus the ability to influence cost and the cost of change across project phases. (Ability to influence cost is greatest at the front end and falls as the project proceeds; cost of change rises correspondingly.)
- Merrow, E. W. Industrial Megaprojects: Concepts, Strategies, and Practices for Success (Wiley, 2011); Independent Project Analysis (IPA) / SPE 153695, Oil and Gas Industry Megaprojects: Our Recent Track Record. (Analysis of 318 industrial megaprojects; front-end loading (FEL) at FID is the most reliable predictor of outcome; where FEL is poor, cost deviations of ~50% or more follow.)