The IEA line
IEA: more than 10 mb/d of Gulf output still shut in as of August. Gulf oil exports ~13 mb/d, nearly half their pre-war level. Insurance, waiting times and the risk premium remain elevated.
Constrained since 28 Feb 202616 September 2026
Chokepoint brief
28 February 2026: Gulf loadings stop. ~20% of seaborne crude and ~20% of seaborne LNG leave the timetable and enter the risk premium. IEA September 2026 still has more than 10 mb/d of Gulf output shut in. This page is that story — not the rest of the atlas.
Gulf shut-in
10 mb/d
~half of pre-war Gulf exports
Exports ~13 mb/d against 25 before the war.
2026 supply
-5.7 mb/d
100.7 mb/d IEA September
Demand -2.5 mb/d. Barrels that cannot sail destroy demand at $100+.
Stocks drawn
507 mb
Since February. The OECD cushion is a calendar, not a replacement for Ras Tanura.
OPEC spare
0.07 mb/d
Shut-in barrels are not spare. They cannot load.
Brent
$107.8
Security premium on the 2026 tape.
Dubai
$105.4
The Asian sour marker. Hormuz cargo.
VLCC AG–CN
$6.4 /bbl
TD3C. Time around the Cape.
JKM
$15.8
$/MMBtu. Qatari trains that cannot sail.
Sequence
The war is a loading programme. These are the marks the physical market still prices.
28 Feb 2026
Gulf loadings halt. ~20% of seaborne oil and LNG is no longer a timetable; it is a risk premium.
Spring 2026
VLCCs stack on the long way around. Red Sea product already had a Houthi problem; Hormuz made it structural.
August 2026
IEA: more than 10 mb/d of Gulf output offline. Exports ~13 mb/d against 25 before the war.
September 2026
The Saudi bypass takes a drone strike. Spare capacity on the IEA sheet is 0.07 mb/d — shut-in barrels are not spare.
Who is offline
Gulf producers are not a single spare-capacity pool. They are six loading programmes and one landlord.
| Producer | Liquids | Net oil | Rent |
|---|---|---|---|
| Saudi Arabia | 11.4 | +7.4 | 24.0% |
| Iraq | 4.4 | +3.5 | 42.0% |
| Iran | 5.2 | +3.2 | 18.0% |
| Kuwait | 2.6 | +2.1 | 38.0% |
| United Arab Emirates | 4.2 | +3.0 | 16.0% |
| Qatar | 1.3 | +1.0 | 28.0% |
| Oman | 1.1 | +0.8 | 22.0% |
Million barrels per day and oil rent as % of GDP, 2025. Click through for the dossier.
The IEA line
IEA: more than 10 mb/d of Gulf output still shut in as of August. Gulf oil exports ~13 mb/d, nearly half their pre-war level. Insurance, waiting times and the risk premium remain elevated.
Constrained since 28 Feb 2026The bypass
A drone strike in mid-September forced a shutdown of the main bypass. Saxo estimated ~4% of global supply at risk while storage fills.
Damaged, partially offlineBypasses
Pipelines and the Cape move some barrels. They do not restore 20% of seaborne oil.
~5 mb/d
Damaged
The main Hormuz bypass. A mid-September drone strike took the line down; storage is filling on the Gulf side.
UAE Murban
Open
Loadings east of the Strait. Some UAE crude skips Hormuz; it does not replace Basra or Ras Tanura.
Time, not barrels
Congested
The long way around. VLCCs stacked on the 12-day detour. Insurance did the rerouting before navies did.
Residual
Thinned
Red Sea risk already pulled product around the Cape. SUMED is not a relief valve while Hormuz is tight.
Who still buys
China, India, Japan, Korea and Europe imported most of what they burned. Then the Strait closed.
73%
The buyer that sets the seaborne tape. ESPO, WTI and Brazil filled the Gulf hole. The SPR is the buffer.
86%
Urals and Atlantic barrels. A Hormuz premium is a current-account event before it is a refining event.
~100%
JKM and Dubai pass straight into industry. The nuclear restart is the only fast LNG hedge.
~98%
Product-export machines that still need sour crude. When Gulf barrels vanish, cracks and diets both move.
75%
US light, Atlantic heavy, products around the Cape. ARA is a residual of whatever can still sail.
See it on the water
The map is the product. Filter crude, toggle 2025 against September 2026, and the hole Hormuz punched is the empty water between Ras Tanura and Ningbo.