On August 11, the European Union's Sentinel-2 satellite captured something that doesn't appear on any AIS tracker, any insurance underwriting database, or any customs manifest. Along a 100-kilometer stretch of water off the coasts of Oman and the UAE, twelve pairs of tankers were pressed hull-to-hull, transferring crude oil between them. Shuttle VLCCs that had crossed the Strait of Hormuz in the dark were handing their cargo to clean vessels that had never entered the insurance exclusion zone.
A virtual port, in open water, visible only from space.
Three Flows
Five months into the Hormuz closure, oil is still moving. But it no longer flows through a single system. It has fractured into three parallel mechanisms, each with a different risk profile, a different threat vector, and a different measurement problem. Map each one and you see the actual supply chain architecture under stress.
Flow 1: Dark Transit
Ships cross the Strait of Hormuz with AIS transponders off. The US Navy coordinates clearances. Iran's IRGC patrols. Both sides maintain a fragile operational fiction that this is manageable.
The numbers tell a different story. Ten-day average transit sits at roughly 11 ships per day — a fraction of pre-closure levels. On August 11, only 6 ships crossed. Bab el-Mandeb traceable transits hit a one-year low: 200 per week, down 38%.
And dark doesn't mean safe.
ADNOC Vessels Attacked — AIS Dark at Time of Strike
Two Abu Dhabi National Oil Company tankers were struck by IRGC drones during outbound Hormuz transit on August 14. Both had transponders off. A bulk carrier was hit the next day. Total ADNOC vessels targeted since the conflict began: 17.
The UAE condemned the attacks as piracy. The attacks confirm what ship operators already knew: AIS-off reduces commercial tracking but does not defeat radar, drone surveillance, or IRGC fast-boat interdiction. Going dark buys ambiguity, not protection.
Flow 2: The Handoff
This is the mechanism that doesn't show up in conventional supply chain analysis.
The mechanism is insurance arbitrage. A shuttle VLCC loads crude at a Gulf terminal, transits Hormuz dark (the high-risk window is roughly 30 minutes), then conducts a ship-to-ship transfer in the Gulf of Oman. The receiving vessel was never inside the insurance exclusion zone. It proceeds to its destination with standard coverage, clean documentation, and no war-risk surcharge.
The shuttle absorbs all the risk. It makes short, repeated dark transits — load, cross, transfer, return. The receiving vessel absorbs none. From an insurance underwriter's perspective, the receiving vessel's cargo originated in the Gulf of Oman, not the Persian Gulf.
On July 18, satellite imagery showed just one STS pair operating — the mechanism had collapsed after a wave of IRGC attacks on transferring vessels. By August 11: twelve pairs, handling at least four VLCCs carrying 8 million barrels. The system adapted, dispersed along 100 km of coastline, and surged.
Flow 3: The Pipeline
Saudi Arabia's East-West Pipeline pushes crude 1,200 km from the Gulf coast to Yanbu on the Red Sea. This was the primary bypass when Hormuz closed — at peak, Yanbu was loading 4.7 million barrels per day, nearly five times its 2025 average.
Then the Houthis declared a blockade.
| Yanbu Metric | Pre-Blockade | Current | Change |
|---|---|---|---|
| Visible crude loadings (bpd) | 4.23M | 2.4–3.0M | -30 to -43% |
| Voyage count (per week) | 35 | 16 | -54% |
| LPG exports (mt/month) | 296,900 | 91,500 | -69% |
| Dark loadings (% of total) | ~5% | ~70% | — |
The visible decline is severe. But the real story is the last row. Seventy percent of Yanbu loadings are now conducted dark — AIS off to avoid Houthi targeting. Tankers conceal their port calls. This means the visible numbers understate actual throughput, but by how much?
Three commercial tracking services give three different answers:
VORTEXA
2.38M
bpd
KPLER
1.78M
bpd
AXSMarine
850k
bpd
A 3:1 ratio between the highest and lowest estimates. The tracker divergence isn't noise — it's the signal. When 70% of loadings go dark, commercial tracking infrastructure breaks. Each provider makes different assumptions about dark vessel behavior, port call inference, and satellite pass timing. The spread between them is itself a measure of how unmeasurable the flow has become.
What Breaks Each Flow
| Flow | Threat | Kill switch |
|---|---|---|
| Dark transit | IRGC drones + fast boats | Sinking, not striking. A hull breach ends the fiction that dark = safe. |
| STS handoff | IRGC targeting STS zone | Attack on transferring vessels. Two hull-to-hull ships are maximally vulnerable — zero maneuverability, flammable cargo exposed. |
| Yanbu pipeline | Houthi drones + missiles | Sustained port infrastructure damage. Jizan already hit (400k bpd offline, restart delayed to Aug 30). Wafa and Daisy tankers attacked near Yanbu. |
Each flow compensates for the others' weakness. When dark transit gets attacked, STS volumes surge. When STS gets attacked, dark transit resumes. When Yanbu stumbles, Gulf-origin flows matter more. But all three are degrading simultaneously. And they share one vulnerability: every barrel that originates inside the Gulf must still cross Hormuz.
What the Deal Doesn't Fix
Iran and Oman are finalizing a "shipping map" — a 60-day temporary arrangement establishing a two-way corridor through Hormuz. Iran controls navigation, monitoring, and security. US and Israeli vessels are banned. Tehran insists the legal regime will not return to its pre-war state.
If the corridor opens, it formalizes Flow 1 — dark transit becomes visible, sanctioned transit. Ships use designated routes, AIS on, under Iranian supervision. This is exactly what Iran wants: institutionalized control over the world's most critical chokepoint.
But the corridor doesn't touch Flow 2 or Flow 3. STS transfers happen outside the strait — they're a response to Hormuz risk, not Hormuz closure. They'll continue as long as war-risk premiums make direct transit uninsurable for some operators. And Yanbu flows face a completely different threat actor — the Houthis don't negotiate with Oman, and the Mecca Defense Pact (Saudi-Pakistan-Turkey) changes their targeting calculus, not their intent.
The measurement problem is the supply chain problem.
When you can't measure what's flowing — because 70% is dark, because STS transfers aren't captured in port data, because tracker providers diverge 3:1 — you can't price the commodity, underwrite the insurance, or assess the exposure. The market trades on guesses dressed as data. Brent at $87 reflects a consensus estimate of how much oil is moving. The actual number is somewhere between AXSMarine's 850,000 bpd and Vortexa's 2.38 million. That range is the real risk premium the market hasn't priced.
Correction
In Post #65 (Every Barrel, July 28), I stated Abqaiq was at full halt with 70% confidence and predicted Yanbu loadings would "collapse further." Yanbu didn't collapse. Five VLCCs were loading on August 3, with throughput averaging above 4 million bpd. I've downgraded full-halt confidence to 40%. Possible explanations: partial Abqaiq restart, Yanbu storage drawdown, or damage less severe than satellite imagery suggested. Satellite smoke plumes do not equal confirmed operational status. Loading data is more reliable than imagery.
Brent $87-88. SPR 298.7M bbl — below the 300M floor for the first time since 1983. Jizan restart delayed to Aug 30 after another Houthi attack Aug 9. Iran-Oman corridor deal imminent but not closed. The three flows mapped here are what's actually keeping barrels moving. Each is fragile. None is permanent. And no one can agree on how much oil any of them carry.