An independent monitor of energy supply risk —
measured, dated, and scored against its own record.

This page tracks a small set of verifiable fuel and inventory indicators, shows where earlier forecasts — including our own — were right and wrong, and describes what would change the assessment. It is deliberately not a countdown. Most of what it tracks is currently unremarkable, and it says so.

— days since the Strait of Hormuz disruption began · Feb 28, 2026

This is an elapsed-duration counter for a dated geopolitical event. It is not a countdown to any outcome, and it does not imply the disruption is ongoing at its original severity. Current transit status is listed in the data table below and is only stated where a reliable source is on hand.

Section 1

Current verified data

Every figure below carries its own source and its own data date. Nothing here is attached to a narrative or a forecast. Where a value cannot currently be sourced, the field says so rather than showing an old number.

Data currency: Checking…

Why several fields are blank. An earlier version of this page carried figures that kept their original styling and prominence long after their sources went stale. Showing an empty, labelled field is more honest than showing a confident old number. Fields fill in as each is wired to a live source.

Historical trends

February 2026 to present

Every point on every chart below is a real weekly EIA observation, read from the authoritative dataset this page is generated from. No chart contains a hand-entered or hardcoded value. Charts are grouped so that only compatible units and magnitudes share an axis.

Section 2

What was forecast — and what actually happened

This section exists because forecasts that quietly disappear are worth less than forecasts that are scored. Both an outside forecast and this project's own May 2026 model are recorded below, with outcomes. The goalposts are not moved.

Outside forecast — "tank bottoms" around July 4, 2026

Jeff Currie, Senior Advisor, Carlyle Group · Bloomberg Television, May 6, 2026 · restated at the UBS Wealth Conference, Singapore, May 25, 2026

What was said

US oil storage would approach "tank bottoms" — the level at which pumps can no longer draw fuel — "somewhere in that July 4th period." This was an outside analyst's forecast. It was not a date independently originated by this project.

What happened

The underlying inventory concern was not imaginary. Cushing, Oklahoma crude inventory did subsequently fall below 20 million barrels for several weeks, and the EIA itself later published an explainer on what "tank bottoms" means — an unusual step that reflects genuine attention to the mechanic.

What did not

The modelled downstream consequence did not occur. There were no widespread pump failures and no national physical supply collapse. Distribution routing, policy intervention, and global market adaptation absorbed the stress before it reached consumers.

Verdict

Partially validated

Correct that a real inventory mechanic was under strain. Incorrect that the strain would transmit into physical shortage. A forecast can identify a true pressure point and still be wrong about the consequence — that is the distinction this scoreboard exists to preserve.

This project's own May 2026 model

Published on the earlier version of this site · scored against EIA weekly data

Forecast

Diesel days of supply would fall through a 25-day "hard allocation" tripwire around June 24, 2026, then through 20 days around September 2, 2026, with regional physical shortages following.

Outcome

Neither threshold was crossed. The series reached its low of 25.5 days on May 1, 2026 — before the forecast date — and then recovered, holding between 27.4 and 30.1 days through the end of July. The chart below is the full record.

Not this project's forecast

The "tank bottoms around July 4" date was Jeff Currie's, published by Carlyle Group and scored separately above. This project reported and tracked that forecast; it did not generate it, and it is not part of this model. It is recorded here only so the two are not confused.

Verdict

Not validated

This model was too bearish on both timing and US physical severity. It over-weighted a single inventory metric and under-weighted the capacity of distribution networks, policy levers, and global markets to adapt. Its own error was the threshold timing above. Separately, this project gave an outside analyst's July 4 forecast more prominence than a third-party projection warranted — but that date was Currie's, and presenting it as this project's own prediction would be inaccurate. The deterministic shortage thesis has been retired as this project's operating view.

US distillate (diesel) days of supply — March 20 to July 31, 2026

Every point is a weekly EIA reading recorded by this project's own data log at the time of publication. The 25-day line is the threshold the May model predicted would be broken in June.

Diesel days of supply 25-day tripwire (forecast to break in June — never broken) Series low: 25.5 days, May 1

What this changes

Because the inventory trend went the other way, every conclusion that depended on it has been withdrawn, not reinterpreted. Specifically: the June 24 and September 2 threshold dates, the prominence this page gave to an outside analyst's July 4 pump-failure forecast, and the claim that the sequence was "playing out exactly as projected" are all gone from this page. They are recorded here as scored forecasts and nowhere else.

Section 3

Current scenarios

Stated as probability bands, not point estimates. Precise-looking percentages on questions like these convey more confidence than anyone actually has, so none are given. No shortage dates are predicted.

Most likely

Managed recovery into a prolonged period of elevated energy costs

Supply continues to route around the disruption, inventories stay within workable ranges, and the durable effect is cost rather than availability. Households and businesses feel this as persistently higher fuel, freight, and goods prices rather than as empty pumps. This is the scenario the currently observable data is most consistent with.

Lower probability

Localized, temporary supply interruptions

Regional disruptions driven by ordinary infrastructure causes — a major refinery outage, a terminal or pipeline problem, or severe weather — producing short, geographically contained availability problems. These occur in normal years too and are not unique to the current geopolitical situation.

Tail risk

Severe nationwide physical shortage or rationing

A genuine national supply failure requiring allocation or rationing. This remains possible and is worth monitoring, which is why this page exists — but it is a low-probability tail, not a base case, and the 2026 record so far is evidence of how much adaptive capacity sits between an inventory squeeze and a consumer shortage. No date is attached to this scenario because no defensible basis for one exists.

Section 4

What would change the assessment

Committed in advance, so that a change of view can be checked against something rather than asserted after the fact. These are directional triggers. Numerical thresholds are given only where the project's own data record supports one.

▼

A sustained multi-week decline in diesel days of supply

Specifically, a decline that breaks below 25.5 days — the low recorded on May 1, 2026 — and stays there for several consecutive weeks rather than rebounding as it did in May and June. A single weekly dip is noise; the 2026 series has produced several.

▲

Continued inventory improvement

Days of supply holding at or above the levels seen through July would further reduce the weight given to the tail scenario.

▼

A severe or extended refinery outage

Particularly one affecting a large facility or a regionally concentrated cluster, where replacement supply cannot be routed in quickly.

◆

A major Gulf hurricane affecting refining infrastructure

Gulf Coast refining and offshore production are seasonally exposed. A significant storm strike is the most common historical cause of genuine, if temporary, US fuel supply disruption.

◆

Material change in Strait of Hormuz shipping — in either direction

A sustained deterioration in transit volumes, or a durable recovery toward pre-disruption levels. Both matter; this page currently asserts neither, because it does not have a verified current figure.

▼

Major military escalation

A significant widening of the conflict, or direct targeting of energy production, export, or transit infrastructure.

◆

Expiry or extension of emergency policy interventions

Fuel waivers, reserve releases, and export policy adjustments have been part of the 2026 picture. Their removal tightens flexibility; their extension preserves it. See the Arizona note below for a current example.

Section 5

Low-regret steps

Only measures that remain sensible whether or not anything further happens. Nothing here depends on a crisis occurring, and nothing here asks anyone to buy ahead of a predicted surge.

Don't routinely run your tank near empty

During any period of fuel-supply uncertainty, keeping a habitual buffer costs nothing and removes the need to make decisions under pressure. This is a habit, not a stockpile.

Keep normal household emergency supplies

The ordinary, widely recommended baseline — water, several days of shelf-stable food, medications, flashlight, batteries, first aid. This is standard emergency preparedness guidance for any hazard and is not specific to energy.

Service vehicles ahead of seasonal demand peaks

Routine maintenance scheduled early avoids competing for service capacity at the busiest times of year. Useful in any year.

Farming households: plan harvest fuel on a normal or slightly early schedule

Where fuel is a scheduled operational input, ordering on a normal cycle with a modest lead time is prudent planning. This is not a recommendation to build unusual reserves.

Winter heating-fuel households: arrange supply on a normal or slightly early schedule

Heating oil and propane customers routinely benefit from arranging delivery before peak season regardless of wider conditions.

Maintain a reasonable financial buffer

General financial resilience against cost increases of any kind. This is a note about ordinary household budgeting, not investment guidance of any sort.

What this page does not recommend: panic buying, large speculative fuel stockpiles, or purchasing any category of goods ahead of an anticipated demand surge. Earlier versions of this page carried urgency-based purchase framing. It has been removed, and its removal is deliberate.

Regional note

Arizona summer fuel requirements and the federal waiver

Included because it is specific, dated, and checkable — and because it is an example of the policy-expiry trigger listed above.

The current EPA nationwide fuel waiver runs through August 28, 2026.
Arizona's state Cleaner Burning Gasoline summer program runs May 1 through September 30.
EPA lists Arizona's maximum vapor pressure as 7.0 psi from May 31 through September 30 under normal state requirements.
If the waiver is not extended, Arizona loses some fuel-supply flexibility while the stricter summer requirements remain in force.
EPA allows qualifying fuel already introduced or certified before waiver expiration to continue being distributed and sold until that volume is depleted — so any effect would phase in rather than switch on at a date.

No specific price impact is published here. This page does not carry a predicted cents-per-gallon figure for waiver expiry, because the size of any effect depends on inventory already in the system, regional supply conditions, and whether the waiver is extended.

Optional tool

Direct fuel cost — illustrative model

Transparent arithmetic on your own direct fuel spend. It multiplies the gallons you actually buy by a price change you choose. It forecasts nothing.

Illustrative model — not a forecast

Gallons you buy per week

Price change per gallon ($)

—

This covers direct fuel purchases only. Indirect effects — freight, groceries, travel, and consumer goods — are real but are not quantified here, because a credible household figure would require assumptions this project cannot currently support with evidence, and naive estimates tend to double-count the same fuel cost across several categories. Those categories are discussed qualitatively rather than priced.

Section 6

About and methodology

What this is

An independent, informational project that monitors a small number of energy supply indicators and publishes them with their sources and dates. It is not affiliated with, and does not speak for, any government agency, data provider, or company named on this page.

How to read the three kinds of statement on this page

Verified data — a measured value with a named source and a data date. Section 1.
Analysis — this project's interpretation of that data. Clearly the author's judgement, and fallible; Section 2 documents a case where it was wrong.
Scenario estimates — qualitative bands describing possible futures. Not predictions, and deliberately not given as precise percentages. Section 3.

On changing the assessment

Forecasts can and do change. When the evidence moves, the stated view is updated and the previous position is kept on the page and scored rather than deleted. The retirement of this project's earlier deterministic shortage thesis, recorded in Section 2, is the first application of that rule.

Data refresh

The verified-data section is refreshed by an automated pull from EIA series and market quotes. That automation updates only the values, data-week labels, and timestamp in Section 1 — it does not touch any analysis or scenario text, so a refreshed number can never make an old conclusion look current. Refresh timing is not guaranteed, which is why each field shows its own age.

Questions, corrections, or source information?

If a figure on this page is wrong, out of date, or missing a source, say so and it will be checked against the primary data and corrected on the record. Questions about method, and pointers to better sources, are equally welcome. This is a correspondence channel for the briefing itself — it is not a mailing list, and nothing is sold here.

An independent monitor of energy supply risk —measured, dated, and scored against its own record.