Cost Benchmarking — BOT & Charter
FPSO BOT & Charter Cost Analysis: Why the Headline Number Isn't a Price
Executive summary. Eleven Brazilian FPSOs in this dataset run on a model where Petrobras never pays a one-time construction price: nine long-term charters (Anna Nery, Maria Quitéria, Anita Garibaldi, Carioca, Almirante Tamandaré, and the Mero field's four units — Guanabara, Sepetiba, Marechal Duque de Caxias, Alexandre de Gusmão) and two BOT contracts under construction (SEAP-I, SEAP-II), plus the P-88 Albacora tender — also structured as BOT — whose bid results aren't covered here (see note below). In both models the contractor keeps the unit and gets paid through years of operating value, bundled into one headline number. That number is not comparable to an EPC price, isn't even reliably comparable between two contracts signed the same day for sister hulls (SEAP-I/SEAP-II), and — once normalized by term — shows a ~36% spread between the two charter contracts with enough data to compare at all. The EPC/EPCC side of the market gets its own page.
Two models, one shared trait: no one-time price
Under a long-term charter, the contractor owns and operates the FPSO for a fixed term — typically 20 to 25 years in this dataset — and never transfers title. Petrobras pays a day rate for the life of the contract; the "value" quoted in press releases is usually the sum of that day rate across the full firm term, not a purchase price.
Under a BOT (Build-Operate-Transfer) contract, the same basic economics apply during the operating period — the contractor builds and runs the unit, paid through the term — but ownership reverts to Petrobras at the end of a defined operating window instead of staying with the contractor indefinitely. It's a hybrid: EPC-style eventual ownership, charter-style financing of the build.
Both models share the trait that actually matters for cost analysis: the headline dollar figure is a bundle of construction cost plus years of operating margin, not a price for the asset alone. Treating it as one is the same category error as comparing a house's sale price to the total rent a tenant will pay over a 25-year lease.
The dataset
| FPSO | Field | Model | Term | Contractor | Value (US$) | Oil bpd | First oil |
|---|---|---|---|---|---|---|---|
| Carioca (MV30) | Sépia | Charter | ≤21yr | MODEC | — | 180,000 | Aug 2021 |
| Guanabara (MV31) | Mero | Charter | 22yr | MODEC | —‖‖ | 180,000 | May 2022 |
| Anna Nery | Marlim | Charter | 25yr | Yinson + Sumitomo | 5.5bn† | 70,000 | May 2023 |
| Anita Garibaldi (MV33) | Marlim | Charter | 25yr | MODEC | — | 80,000 | Aug 2023 |
| Sepetiba | Mero | Charter | 22.5yr | SBM Offshore | —‖ | 180,000 | Jan 2024 |
| Maria Quitéria | Jubarte | Charter | 22.5yr | Yinson | 5.2bn† | 100,000 | Oct 2024 |
| Marechal Duque de Caxias | Mero | Charter | 22.5yr | MISC Berhad | —‖‖ | 180,000 | Oct 2024 |
| Almirante Tamandaré | Búzios | Charter | — | SBM Offshore | — | 225,000 | Feb 2025 |
| Alexandre de Gusmão | Mero | Charter | 22.5yr | SBM Offshore | —‖ | 180,000 | May 2025 |
| SEAP-II (P-87) | Sergipe Águas Profundas | BOT | — | SBM Offshore | 4.1bn | 120,000 | 2030 |
| SEAP-I (P-81) | Sergipe Águas Profundas | BOT | — | SBM Offshore | 4.3bn | 120,000 | 2031 |
Sister hulls, same day, different price
SEAP-I and SEAP-II are as close to identical twins as this dataset gets: the same SBM Offshore Fast4Ward hull design (the 11th and 12th multi-purpose floater hulls in the program), signed the same day, for neighboring fields in the same Sergipe Águas Profundas project. SEAP-II is publicly reported with higher gas-handling capacity (425 MMscfd versus 355 MMscfd) than SEAP-I, yet its headline value is roughly $200 million lower — $4.1bn versus $4.3bn.
That's a real, sourced divergence between two contracts that share a hull design, a buyer, a contractor, and a signing date — the tightest possible comparison available anywhere in the BOT dataset. If two sister hulls signed on the same day can move in opposite directions from what their specs would predict, a $/bpd or $/ton figure pulled from a single BOT or charter contract in isolation should be treated as one data point, not a benchmark.
Normalizing by term: the real charter spread
Anna Nery and Maria Quitéria are the only two charter contracts in this dataset with both a published total value and a published capacity — but their terms differ (25 years versus 22.5 years), so comparing the raw totals, or even the raw $/bpd, still mixes two different contract lengths into one number. Dividing $/bpd by the term in years isolates what each contract actually charges per barrel of capacity, per year of commitment:
| FPSO | Value (US$) | Oil bpd | Term | $/bpd | $/bpd/year |
|---|---|---|---|---|---|
| Maria Quitéria | 5.2bn | 100,000 | 22.5yr | $52,000 | $2,311 |
| Anna Nery | 5.5bn | 70,000 | 25yr | $78,571 | $3,143 |
Even normalized to an annual basis, Anna Nery charges ~36% more per bpd per year than Maria Quitéria. That's a bigger gap than either the $/ton or $/bpd spread found on the EPC side — on a sample of just two. It's a reminder that charter pricing carries variables EPC pricing doesn't: financing structure, the contractor's cost of capital over a quarter-century commitment, and field-specific risk (Anna Nery and Maria Quitéria are both mature-field revitalizations, but in different basins with different reservoir profiles) all show up in the day rate in ways a single $/bpd number can't separate out.
Financing isn't the same number as contract value. Two of the four Mero-field charters — Sepetiba and Alexandre de Gusmão — only have a public project financing figure (US$1.6bn and US$1.615bn), not a total charter value. Financing is the debt SBM Offshore raised to build the unit; the charter value would be the full 22.5-year revenue stream Petrobras pays on top of that, which is a separate and necessarily larger number that neither company has disclosed. Quoting the financing figure as if it were "the cost of the FPSO" — easy to do, since both are headline dollar figures attached to the same project — is a third version of the same error this whole page is about: two numbers that look interchangeable and aren't.
Why Petrobras is leaning back toward BOT
Petrobras has been publicly signaling a shift toward the BOT model for future proprietary units, explicitly framed around controlling costs after a run of expensive long-term charters. The logic is straightforward once the two models are laid out side by side: a charter's headline value is, structurally, a multi-decade revenue stream for the contractor with no cap on how long Petrobras keeps paying for an asset it will never own. A BOT contract caps that exposure — operating payments stop at a defined transfer date, after which the unit becomes Petrobras' own asset with no further day-rate obligation.
That shift is also why the contract-model mismatch this whole page is about matters beyond methodology. If BOT becomes the default for new proprietary units going forward, more of the market's future headline numbers will carry this same "bundled value, not a price" caveat — and fewer will be directly comparable to the EPC contracts that dominate the historical record.
A note on P-88 Albacora. Albacora is itself structured as a BOT contract — 120,000 bpd, a Macaé (Porto de Imbetiba) offshore base, and a 20-year design life, per the public tender documents. That's as far as this page goes: the tender's bid results are still being finalized, and I'm a party to that process through my work with the winning consortium's O&M partner, so publishing a cost comparison here — while it's still open — isn't something I'll do. The specs above are the tender's own public terms, not commentary on any bid.
I work on exactly this kind of cost and bid analysis day to day, on the Brazilian FPSO/BOT market. If you're evaluating a bid, a benchmark, or a contract structure and want a second set of eyes, get in touch.