Iran–Gulf developments — 4 October 2026
The strait carried more this week and was struck more. Kpler put September crude exports from the Gulf, excluding Iran, back at the pre-war average, but by a changed route: three-fifths across the strait on shuttle tankers and transfers off Sohar and Fujairah, the rest from Fujairah and the Red Sea. LNG had its busiest month of the war, at nineteen to twenty-one cargoes, still a fraction of pre-war. US-facilitated transits ran near thirty a day, while independent counts of other traffic stayed in single digits. Between 28 September and 2 October six tankers were hit by unknown projectiles inside the strait, two Kuwaiti and three Emirati-managed among them. Nobody claimed them. No cover was found written, and no navigation régime was agreed. The east–west pipeline is reported pumping and Yanbu loading again, and the operator still has not said so. On the Red Sea front, the coalition attributed a drone strike on a Medina power station to the Houthis, who denied it. Government forces struck Sanaa, and Reuters reported a Saudi offensive to retake the Bab el-Mandeb shore in planning, with no date. Through Qatar, Washington answered Iran's seven-day plan. The dispute is over the order of steps.
The developments
Kpler reported on 29 September that crude exports from the Gulf, excluding Iran, were at least 16.5 million barrels a day in September, the pre-war average. About 60 per cent (9.9 million) physically crossed the strait, mostly on shuttle tankers; 23 per cent loaded on the Gulf of Oman coast, mainly at Fujairah; and 17 per cent left through the Red Sea. Before the war 83 per cent crossed the strait. Kpler states its figures are minimum confirmed volumes. S&P Global counted 19 LNG cargoes through the strait in September, 13 Qatari and 6 Emirati, and Kpler counted 21. Both are the highest month of the war, against 15 in June and no visible Qatar-linked transit in August. Many carriers crossed with transponders off. The Joint Maritime Information Center recorded 113 US-facilitated transits in the four days to 30 September, while independent tracking of other cargo vessels showed single digits each way. Its 2025 baseline is about 138 a day. UKMTO and the trade press reported six tankers struck by unknown projectiles inside the strait. One was on 28 September, three on 29 September, one on 1 October (crew evacuated) and one on 2 October, that last resting on a security firm and trade press only. Security firms and the IMO's list named two Kuwait Oil Tanker Company VLCCs and three ADNOC- or Anglo-Eastern-managed tankers. No party claimed any of the strikes. Iran's Strait Authority said the ships targeted included vessels on its non-compliant list, which reached 82 and now carries warnings to charterers. No war-risk cover was found written, no Joint War Committee change was identified, and the Salalah ministerial had not reconvened.
Why it matters. Volume and access have separated. Crude is leaving the Gulf at the pre-war rate because the route was rebuilt around the strait's risk. State-company shuttles run short legs to transfer points outside it, under US facilitation, and the rest goes by the bypasses. That is what the independent count shows: almost nothing else moves. LNG is the cleaner test, because it has no bypass. It rose, but to about a fifth of pre-war. The strikes fell on exactly the cargo that makes up the recovery, national oil company tankers, and the week's six are the most in the strait in any window this module has recorded since August. Commercial access did not move, and legal access hardened as the Strait Authority extended its pressure to service providers. The strait is used more and is still not open, and nothing here says when a market will return to it. The item does not establish who fired: no party claimed a strike, and no authority attributed one.
- Update 101 to JMIC Advisory Note: Regional Tension — Impact on Maritime Security
- Explainer: How Mideast Gulf crude exports returned to pre-war levels
- LNG shipments through Strait of Hormuz hit post-war high in September
- Defying conflict, Qatar-linked LNG tankers return to the Strait of Hormuz
- Four vessels struck in Hormuz in 24 hours
- Three Oil Tankers Come Under Fire in Strait of Hormuz
- Two More Tankers in Hormuz Report Strikes, Bringing Total to Six in a Week
Reuters reported on 29 September, on trade sources and shipping data, that tanker loadings at Yanbu had resumed after the pipeline restart. An Asian refining source said Aramco had notified customers of its October Yanbu programme on 28 September. Bloomberg reported overseas shipments restarted after a 17-day halt. TankerTrackers.com read European Space Agency imagery of 27 September as showing nearly 10 million barrels loading at Yanbu and Al Muajjiz. Throughput on the line was put at about 2 million barrels a day by an industry source and 2.65 million by Kpler, which expected 3 to 4 million within days. On 1 October the Joint Maritime Information Center called the restart unconfirmed and said regular Yanbu loadings were not independently confirmed. Aramco did not respond to requests for comment, and no ministry or operator statement of a restart, a rate or a damage assessment had been published by 3 October.
Why it matters. The Red Sea outlet is reported working again, after an interruption of about seventeen days, at roughly a third of the line's nameplate. Every figure for it belongs to someone who does not run it. The module keeps the two layers apart. The standing question's bypass-loadings reading moves, because loadings are reported resumed on several independent kinds of evidence. The pipeline-restart observable does not, for the third review running, because it asks the ministry or the operator to say so and neither has. What this item does not establish is the rate. The trackers' estimates differ from one another, the agency that tracks the water would not yet confirm regular loadings, and the April and September damage have still not been described by anyone accountable for the line.
Riyadh is reported planning to retake the Bab el-Mandeb shore, and planning is not an offensive
Reuters reported on 2 October, citing six people including Gulf and Yemeni officials and Western diplomats, that Saudi Arabia is planning an offensive against the Houthi movement. It would be led by Yemeni forces overseen by Riyadh and supported by Saudi air strikes, and is "expected to be launched in the coming weeks". Two options are under consideration: a narrow push around Bab el-Mandeb, or that push combined with attacks on several fronts inland. The sources' estimates of timing ran from within a week to after the United States midterm elections. Western diplomats and a Gulf official said Riyadh has concluded it cannot enter new talks while the movement holds the strait. Reuters reported United States intelligence, advisers and targeting information in support of Saudi operations, with no direct US participation. Axios reported the same day that Washington will not join "for now". Pakistani officials, unnamed, described equipment flown to Aden and Pakistani advisers working with Saudi counterparts. Drones Saudi Arabia bought from Türkiye before the war are reported in use. Saudi, Yemeni, Houthi and US spokespeople did not respond. The Joint Maritime Information Center reported no interference with shipping at the strait in its period to 1 October.
Why it matters. This is the first credible report that Riyadh means to contest the shore the southern strait turns on, rather than absorb the movement's hold on it. It also puts the strait into the Saudi-Yemeni negotiation: no talks while the movement holds the coast. It is reporting of intent, without a date, and the module records neither an offensive nor its start. The Bab el-Mandeb question's coalition-operation observable is not met by a planned ground campaign. The Makkah Agreement did not move either. The Pakistani support is described by unnamed officials and attributed by no government to the Agreement, the 2025 bilateral instrument or anything else, and the attribution is what that question asks for. What would separate planning from a campaign is an acknowledged Saudi or coalition operation with a stated objective on the coast.
On 1 October the coalition's spokesman said a drone struck the Taibah electricity distribution station in Medina, which supplies the Prophet's Mosque, at 05:11 on 29 September. He said one transformer was taken out of service without affecting the wider network, and he attributed the attack to the Houthi movement on the evidence of the drone's wreckage. The movement denied involvement. Türkiye, Qatar, the UAE and the Arab League condemned the attack. That evening AFP heard loud explosions in Sanaa, and the Yemeni government's military said it had struck Houthi targets in Sanaa, Saada and Taiz. The movement's news agency called it Saudi aggression. On 3 October a Reuters witness saw a large plume of smoke and fire near an Aramco facility in Riyadh, and satellite fire detection showed heat at the refinery. The movement claimed a missile and drone strike on Aramco in the capital. The coalition's spokesman called the claims misleading, and neither Saudi authorities nor Aramco confirmed a strike or gave a cause.
Why it matters. Until this window, almost every record on this front was the movement firing and Saudi Arabia intercepting. Now the Saudi side is acknowledged on the record too: strikes in Yemen are being announced by the forces Riyadh backs. That is what the standing question reads, and its assessment has been advanced to record a two-sided front. The reach has also extended, on the coalition's account, to a second holy city's utilities. That account is disputed by the movement, and Vigil records both. The Riyadh fire is not counted as a strike. It is a fire with a claim attached that the coalition rejects, at a site the module does not publish as a marker. None of this is the reported offensive, and none of it is a channel act. No truce instrument exists, and no mediated round has been announced.
Washington answered Iran's plan through Doha, and what separates them is the order of steps
On 28 September Iran's foreign minister said Iran's plan had been passed to Washington through the Qatari mediator, and that the strait would open only if conditions emphasised by the Leader were implemented. The same day the United States told Iran's delegation to leave New York, according to Axios as carried by Gulf News; it was already scheduled to leave that night. On 29 September Qatari mediators in Doha passed the foreign minister Washington's feedback on a seven-day trust-building plan, which aims to return the two sides to an enhanced version of the June memorandum and includes steps on the nuclear programme. A non-Iranian official briefed on the talks told Reuters the disagreement "centres on the sequencing of the steps rather than the components". Iran's government spokeswoman said the US proposal was presented to the cabinet on 30 September. The president said on 1 October that he would not accept proposals he would not have approved a year ago, and left strikes after the 3 November election open. The Treasury designated Iranian automotive, rail, manufacturing and metals companies the same day. No reply from Tehran, next round or prisoner or funds exchange was found by 3 October.
Why it matters. The channel moved one step: a written US response was delivered and is under Iranian consideration. A source close to the talks names the obstacle as the order of steps rather than their content. That matters for the strait, which is Iran's main bargaining asset under this plan. A dispute about sequencing is narrower than one about terms, and it is also the kind that has stalled earlier arrangements in this file. Nothing is in force. There is no framework, no schedule and no text, and the pressure track hardened in the same days: the delegation ordered out, new designations, and strikes after the midterms openly kept on the table. What would separate a channel from a process is a reply with a date for the first step.
Where this fits
This is one conflict's dated review for the week. The cross-conflict edition for the same week carries the most consequential findings from every conflict Vigil covers. The standing picture this review updates is the Iran–Gulf coverage. How Vigil grades, dates and sources its claims is set out in the methodology, and anything later found wrong is recorded in the corrections log.
By email Get Vigil’s Weekly Briefs and new Analytical Notes by email. What Vigil Dispatch sends
← Previous Iran–Gulf brief — 2026-09-27Every Iran–Gulf review All weekly briefs