Oil & Gas Projects Under Evaluation and Design: A Segment-Wise Read of 2026 (Upstream to Petchem)
Data caveat. GEM has not published a 2026 global pre-construction tally in the material I could reach, and S&P Global's project-level pre-FID data is paywalled. Every number below is attributed to its source and year. Where a figure is old (GEM's pipeline numbers date from 2022, its LNG numbers from early 2024), the chart says so. Read the shapes, not the decimals.
The big picture
The pipeline is bifurcating. Capital is flowing to low-breakeven offshore and to US-centred LNG, while a very large stock of proposed gas infrastructure sits unsanctioned, and Middle East and pipeline projects face delay and stranded-asset questions. Wood Mackenzie counts 23 projects sanctioned in 2026 so far, carrying about $96bn of commitments, with up to 10 more (about $70bn) possible but described as unlikely to all land.
Segment 1: Offshore and deepwater (FPSO-led)
This is the healthiest segment. ADI Analytics reports offshore FIDs of about $164bn, the highest in over a decade, and Wood Mackenzie notes FPSO projects dominated Q2 committed investment. The economics explain it: deepwater breakeven is roughly $40/bbl against about $65/bbl for new US shale wells, which mattered when ADI’s January 2026 outlook saw Brent at $55-60 on a 2-4.5 MMbpd surplus.
Segment 2: LNG liquefaction
LNG is where the volume is, and also where the slowdown is. 2025 was a record: 9 projects, 72 Mtpa sanctioned (Wood Mackenzie), over 100 bcm/yr by the IEA's count. For 2026, Wood Mackenzie expects just 4-5 FIDs, and every 2026 FID so far is American (over 31 bcm/yr). Behind them sits an enormous unsanctioned stock: roughly 648 Mtpa proposed in North America alone, 170 in Russia and 133 in Africa (IGU, 2024 data). Whether even a fraction is built depends on contracting in a softer market.
Segment 3: Gas pipelines and import terminals
GEM's last global pipeline census (2022) found 122,500 km in pre-construction against 70,900 km being built, about $486bn of capex flagged as stranded-asset risk. The pre-construction stock was nearly 1.7 times what was physically under way. The regional picture is more concrete: GEM's Asia Gas Tracker says India has the world's third-largest buildout, with nearly three-quarters already under construction, so there the pipeline is converting to steel rather than lingering in design.
Segment 4: Onshore upstream and new fields
The price shock did not set off an onshore boom, and that is the story of this segment. Brent averaged about $114 a barrel in September and WTI averaged $98.70 while the Dallas Fed was surveying producers that month, yet executives said they still budget off $65–70 because the futures curve is steeply backwardated. Asked in March, firms said they needed $66 on average to drill a new well profitably ($67 in the Permian, $59 for large firms). Spot prices clear every one of those breakevens. What holds drilling back is doubt that the prices will last: the same respondents expect WTI at $88 by year-end and $79 two years out.
- Planning price: $55–60 for 2026, assumed a year earlier (Spears & Associates); now $65–70 (Dallas Fed survey comments, September 2026).
- New-well breakeven (WTI): about $65 for new US shale wells (ADI, January); $66 average and $67 in the Permian (Dallas Fed, March 2026).
- US Lower 48 rigs (annual average): 561 in 2025, about 560 in 2026, 605 forecast for 2027 (Spears).
- US crude output: 13.66 mb/d in 2025. EIA’s 2026 forecast rose from 13.51 mb/d in April to 13.87 mb/d in October; 14.30 mb/d is forecast for 2027.
- Cash flow reinvested: about 50% across operators (Spears); about 30–32% at Diamondback.
So the response came through output, not new rigs. EIA raised its 2026 US production forecast in every monthly outlook from April to October, mostly on the Permian. Public companies held their rig counts roughly flat, and Spears says the growth came from privately held operators. Diamondback, a large Permian producer, is typical. It raised its 2026 oil guidance to about 524,000 b/d and its capital budget only from about $3.9bn to $3.95–4.0bn, and is using the extra cash for dividends, buybacks and debt. In the Dallas Fed’s third-quarter survey the E&P capital-spending index cooled from 42.7 to 35.3. Respondents cited rising finding and lease-operating costs and, for independents, tighter bank credit.
Outside the US the shock has been a disruption, not a windfall. Rig work in the Gulf has been interrupted: H&P cut its planned Saudi rig reactivations for the year from seven to five. Saudi Arabia and the UAE, which each have more than 1 mb/d of export routes that avoid Hormuz, have kept drilling, and ADNOC Gas sanctioned Phases 2 and 3 of its Rich Gas Development in August. Argentina’s investment regime has drawn about $60bn of proposed upstream capital and more than 3,000 well commitments. Westwood expects the world’s average onshore rig count to rise about 11% between 2026 and 2031, from 4,021 to 4,445, and India’s from 134 to 154.
New-field approvals have kept running regardless. Global Energy Monitor’s March 2026 extraction-tracker release counts at least 10 billion boe approved for development in 2025, led by volume in Kuwait, Iran, Brazil and the US, against about 8 billion boe across at least 20 fields in 2023. Russia, Norway, Nigeria, Australia and the US hold 34% of fields not yet producing. Southeast Asia remains a regional hotspot: GEM flagged a decade-high number of gas decisions for 2025.
The binding constraint onshore has moved from price to conviction. On January’s $55–60 outlook, shale at a $65 breakeven was the marginal barrel. At $90–110 spot it is comfortably in the money, but operators are budgeting as if prices will fall back to about $65–70. That is a floor about $10 higher than a year ago, and still within a few dollars of their breakevens. Onshore design work is therefore growing slowly from a higher base, not being deferred. The test is whether 2027 budgets, set in February, move their planning price up.
Segment 5: Oil pipelines
GEM's 2022 oil pipeline report had the US ahead of every other country in new capacity by length, with pre-construction projects adding about 21.8 mb/d. I found no 2026 update, so treat this segment as the least evidenced here.
Segment 6: Refining
Refining is a growth story with an expiry date. BloombergNEF projects net capacity growth of about 4.2 mb/d over 2026-2030, with retirements (mainly Europe and North America) taking roughly 0.2 mb/d out. About half of the planned additions are already operating or under construction; the 2028-2030 tranche is at earlier stages and carries more delay or cancellation risk. OPEC's regional split puts 2.8 mb/d in Asia-Pacific, 1.0 in the Middle East and 0.8 in Africa. An earlier IEA outlook called this wave of growth likely to be the last. Near term, the IEA's May 2026 Oil Market Report expected 2026 refinery throughput to fall by 1.6 mb/d to 82.3 mb/d on infrastructure damage and supply disruption, a reminder that projects in design compete with a shaky operating base.
Segment 7: LPG
LPG is the clearest case of midstream capex chasing associated-gas growth. US Gulf Coast export capacity is forecast to rise from about 60 to 100 mtpa by 2029. Projects in design or build include Enterprise's Houston Ship Channel expansion (+300 kb/d, a 36% lift to 1.135 mb/d dedicated, due late 2026), Targa's Galena Park (444 to about 625 kb/d by Q3 2027), and a new 400 kb/d Texas City terminal from Oneok and MPLX. Canada shows the risk: AltaGas's REEF terminal slipped to March 2027 with costs up 12% to about CAD 1.5bn. I found no India demand or import data this round, so the demand side is unaddressed here.
Segment 8: Petrochemicals
This is the weakest segment on the margin. China added more than 10 Mt/y of ethylene capacity in 2024-25, Asian olefin margins have been negative for years, and some Northeast Asian crackers have shut. Even so, SunSirs counts over 13 Mt/y of world-scale cracker capacity in late construction (e.g. Huajin Aramco, Fujian Gulang), against only about 6.87 Mt/y of estimated capacity reductions (CGS International), so supply keeps outrunning demand. Mysteel expects an H1 2026 lull in new polyethylene capacity, with most arriving in H2. Evidence points to delays rather than cancellations, and CERA does not expect margins to turn positive until after 2030. The design-phase signal is that new Chinese greenfield investment slows after 2029, while integrated refinery-petchem complexes remain the preferred form in Asia and the Middle East.
Segment scorecard
| Segment | Pipeline stage signal | Direction in 2026 | Evidence quality |
|---|---|---|---|
| Refining | ~4.2 mb/d net adds to 2030; later tranche early-stage | Asia/Africa/ME growth; West closing | Medium (BNEF, OPEC) |
| LPG export | Multiple Gulf Coast expansions to 2027; Canada slipping | Strong, US-centred | Medium (company and trade press) |
| Petrochemicals | Late-stage build still landing into oversupply | Weak margins; delays | Medium (SunSirs, S&P, Mysteel) |
| Offshore / FPSO | Heavy sanctioning | Strong | Medium (ADI, WoodMac) |
| LNG liquefaction | Record 2025, slowdown, huge pre-FID backlog | Slowing; US-centred | Medium-high |
| Gas pipelines / terminals | Pre-construction stock exceeds build | Mixed; India converting | Low (2022 global, 2026 regional) |
| Onshore / shale | In the money at spot; budgeted at $65-70 | Disciplined: output up, rigs flat | Medium (Dallas Fed, EIA, Spears) |
| Oil pipelines | US-dominated proposals | Unknown | Low (2022) |
| Middle East | Delays flagged | Delayed | Low-medium |
What to watch
- LNG contracting. The 4-5 FID forecast versus 9 last year is the clearest slowdown signal.
- Planning price versus spot. Shale breakevens (about $66) sit far below spot, but budgets assume $65–70. If Middle East supply returns and prices fall towards that floor, anything with a breakeven near $65 is back at risk of deferral.
- Middle East timing. Delays shift capex into 2027.
- Downstream timing. Watch whether 2028-2030 refinery tranches slip, whether LPG terminals hit 2027 dates, and whether Chinese crackers start in H2 2026 as scheduled.
- Fresh GEM data. A new GOGET release or gas tracker update would replace the stale numbers above.
