Three firms measured oil loadings at Yanbu in the same week. Vortexa reported 2.38 million barrels per day. Kpler reported 1.78 million. AXSMarine reported 850,000. Same port. Same ships. Same week. Divergence: 3 to 1.
This is not a data discrepancy. It is the collapse of the measurement system itself.
How Oil Gets Counted
Global oil tracking rests on a single technology: the Automatic Identification System. AIS was designed in the 1990s as a collision-avoidance tool — ships broadcasting position, speed, and heading so they wouldn't hit each other. Over the next two decades, commodity tracking firms repurposed these signals into the foundation of energy market intelligence. Vortexa, Kpler, AXSMarine, Windward, and others built their platforms on one assumption: ships broadcast.
That assumption died when broadcasting became a targeting vector.
The sequence
1. AIS built for navigation safety — ships voluntarily broadcast identity and position
2. Commodity firms repurpose AIS data into oil flow measurement and price discovery
3. IEA, EIA, OPEC build supply/demand models on tracker estimates
4. Financial markets price oil based on these models
5. War begins. UKMTO: “Active AIS may be a targeting factor”
6. Ships go dark. Extended AIS blackouts — days, not hours
7. Tracker divergence follows. Same port, 3:1 spread
8. Supply estimates become confidence intervals, not measurements
The mechanism is simple. When ships go dark, trackers fall back on satellite imagery — counting blobs at berths, estimating draft changes, inferring cargo volumes from hull waterlines in low-resolution photographs. Some firms use synthetic aperture radar. Some cross-reference port agent reports. Each uses different methodology, different satellite passes, different interpolation models. Without AIS as the common signal, they diverge.
The Scale
This is not a fringe phenomenon affecting a few sanctioned tankers. It has consumed the majority of Gulf oil flows.
| Metric | Value | Source |
|---|---|---|
| Hormuz transits running dark | 57% | Windward |
| Peak dark transit rate | 65.2% | Windward, May |
| Saudi west coast loadings dark since Jul 23 | 70% | Kpler |
| Yanbu liftings dark (latest week) | 100% | Vortexa analyst |
| Ships off Oman coast for STS | ~150 | vs 40 in January |
| Dark fleet vessels (global) | 1,529 | TankerTrackers |
| Yanbu voyage count decline | -54% | 35 → 16/week |
On August 19, Windward's North Hormuz Corridor tracker detected 16 contacts crossing the strait in 24 hours. Nine were dark. Five had no AIS, no IMO number — detected only by satellite imagery. These are not sanctioned Iranian tankers with known deceptive practices. These are commercial vessels hauling Gulf crude that have decided broadcasting their position is not worth the risk of an IRGC cruise missile.
What Breaks
When a Vortexa analyst says “all Yanbu liftings conducted dark” and Kpler simultaneously reports 70% dark for the same region, the problem is not that one firm is wrong. The problem is that neither firm can verify the other. They are both estimating, using different satellite imagery, different algorithms, and different assumptions about what a hull shadow means. The common reference signal — AIS — is gone.
This propagates upward through the entire information architecture of oil markets:
The IEA's August report called Q3 2026 deficit 1.8 million barrels per day — doubled from their prior estimate. But that number rests on tracker data that diverges 3:1 at the source. The IEA acknowledged the problem directly: dark transit is “complicating the tracking of oil and gas flows and obscuring visibility.” The deficit could be 1.2 million barrels. It could be 2.8 million. The measurement system cannot distinguish between these outcomes.
This is not uncertainty at the margin. This is uncertainty at the center of the market.
The Yanbu Problem
Yanbu is where measurement collapse is most visible, because Yanbu is where the bypass was supposed to be measurable.
Before the Houthi blockade (pre-July 20), Yanbu was loading 4.23 million barrels per day. Voyage counts, AIS tracks, and port agent reports all converged. The number was reliable. After the blockade, ships started going dark to avoid Houthi targeting in Bab el-Mandeb. Within three weeks, all three major trackers had diverged:
Vortexa
2.38
mb/d
Kpler
1.78
mb/d
AXSMarine
0.85
mb/d
Same port. Same week. Mid-August 2026.
The gap between the highest and lowest estimate is 1.53 million barrels per day. For context, that gap alone is larger than Libya's total oil production. It is larger than the entire output of the North Sea Forties system that anchors Brent pricing. It is not a rounding error. It is a measurement system that has lost its common reference.
Kpler reports 70% of Saudi west coast loadings have been dark since July 23. A Vortexa analyst stated that in the latest week, “all Yanbu liftings conducted dark.” When every loading is dark, every volume estimate is a satellite-derived inference. The measurement system has not degraded — it has been replaced by something fundamentally different.
Why the Market Stays Calm
Brent is at $93.86. Not $60 (the supply is fine). Not $150 (the supply has collapsed). Somewhere in between — pricing a world where oil is probably still flowing, probably in reduced volumes, but nobody can say how much. The market has substituted satellite blob counts for AIS-derived flow data, and is treating the result as adequate.
Fortune reported on August 16 that dark flows are “keeping global prices in check” — the market is assuming oil is still moving based on satellite imagery showing ships at berths. This is not irrational. Ships are loading. Crude is moving. But the precision of how much has collapsed from barrel-level tracking to hull-shadow inference.
“Structural breakdown in the information architecture that underpins global energy markets.”
— OilPrice.com, August 2026
The risk is asymmetric. If actual flows are higher than satellite estimates suggest, prices are slightly overvalued — manageable. If actual flows are lower — if the IEA's 1.8 mb/d deficit is an undercount — the market is sitting on a data void that could reprice violently when physical delivery shortfalls materialize. Nobody knows which scenario is true, because the measurement system cannot tell them.
Monday
On August 25, Treasury Secretary Bessent will hold a press conference to unveil what he calls “the toughest sanctions in history” against Iran. He told CNBC it would be a “one-two punch” — the naval blockade plus financial warfare. He told international partners: “You’re either with us or against us.”
The targets he outlined: shadow banking networks, digital-asset exchanges, and shipping networks tied to sanctioned oil sales.
This is the supply chain intersection that makes measurement collapse matter now. The dark fleet — the 1,529 vessels running with transponders off, conducting ship-to-ship transfers off Oman, loading at Yanbu without AIS — is simultaneously:
The mechanism keeping oil flowing. Without dark transit, the remaining bypass infrastructure delivers less. Dark ships carry the barrels that keep Brent below $100.
The mechanism breaking measurement. Without AIS, nobody knows how many barrels those ships carry. The 3:1 tracker divergence exists because these ships exist.
The target Bessent just named. If Treasury sanctions the intermediaries — the STS operators, the ship managers, the digital-exchange payment rails — it degrades both the oil flow and the measurement of that flow simultaneously.
Sanctions on dark fleet infrastructure don't just reduce supply. They reduce the ability to measure the supply they reduce. The information void gets deeper at the exact moment the physical constraint gets tighter.
What This Means for the Map
I have spent 68 posts mapping supply chains — tracing who supplies whom, where the chokepoints are, how disruption flows through dependency graphs. This post is different. This is about what happens when the measurement layer underneath the map breaks.
Supply chains have three layers: physical infrastructure (ships, pipelines, ports), measurement infrastructure (AIS, trackers, port agents, customs data), and financial infrastructure (contracts, insurance, pricing benchmarks). I have spent six months mapping the first layer — how Hormuz closure cascades through bypasses, how Houthi blockade degrades Yanbu, how ADCOP became the last thread.
But the second layer — measurement — has been failing in parallel, and I have not given it the attention it deserves. If you cannot measure a supply chain, you cannot map it. If you cannot map it, you cannot price it. The 3:1 divergence at Yanbu is not a data quality issue. It is a structural failure in the information infrastructure that connects physical barrels to financial markets.
The oil is still flowing. The ships are still loading. The bypass — degraded, attacked, running dark — still carries barrels from the Gulf to Asia. But the system that counts those barrels, that converts hull shadows into supply estimates, that feeds those estimates into IEA reports and trading algorithms and price discovery — that system is broken. We are pricing 20% of the world's oil based on satellite photographs and informed guessing.
On Monday, Bessent plans to target the ships that produce the guessing. Less oil and less visibility, at the same time. The fog thickens.
Brent $93.86. Day 174 of Hormuz closure. Yanbu 100% dark.