1. The short answer
A barrel count alone does not tell you whether supply is tight, because it says nothing about how fast that stock is being drawn down; days of supply does, by dividing the stock by average daily demand. US distillate stocks, diesel, heating oil and jet fuel come from the same refinery cut, stood at 107.9 million barrels in the week to 11 September 2026, 13% below the five-year average, according to the EIA's Weekly Petroleum Status Report. That is a real, sourced number, and it still does not answer the question "for how many days." Our diesel shortage entry covers the wider condition a falling stock like this one often signals; this page covers the day-count arithmetic underneath it.
2. How the ratio is built, and why refining and reporting cadence matter
Days of supply is stock divided by average daily demand, but which stock and which demand figure an agency uses changes the answer, and agencies do not all use the same ones. The EIA calculates a version for the United States from weekly and monthly ending stocks divided by "product supplied," its closest proxy for demand, publishing it in the Short-Term Energy Outlook and in Weekly Petroleum Status Report supplementary tables. The IEA instead reports "days of forward cover" for OECD countries, dividing commercial stocks by the demand it expects over the following months rather than demand already used, in its monthly Oil Market Report.
Reporting cadence differs too: the EIA publishes weekly, while European gasoil stocks at the Amsterdam-Rotterdam-Antwerp hub, covered in our ARA barge price entry, are reported by Insights Global only once a week, on Thursdays. How much distillate gets produced in the first place also feeds the stock side of the ratio, a separate question our refinery utilisation entry covers on its own terms. None of these differences make one agency's number wrong, but they mean a US days-of-supply figure and a European one are rarely directly comparable.
3. The stock numbers that exist, and the days figure that does not
Here is what is actually documented, each figure with its own date and source, followed by an honest gap: no current days-of-supply number for the US, the OECD or Europe as a whole turned up in a dated, publicly accessible form as of 18 September 2026. The EIA's own dedicated data series for this exact metric, pet_stoc_days_dc_NUS_num_w, returned a 404 error on retrieval. Until a verifiable figure surfaces, this page works with the barrel counts below rather than estimate a days figure that cannot be checked.
| Indicator | Value | As of | Source |
|---|---|---|---|
| US distillate stocks | 107.9 million barrels, 13% below the five-year average | Week to 11 September 2026 | EIA, Weekly Petroleum Status Report |
| ARA gasoil stocks | 11.90 million barrels, a four-year low | End of August 2026 | Reuters/Insights Global via IndexBox |
| ARA gasoil stocks, mid-September | 12.08 million barrels, "practically unchanged" | September 2026, exact date not verifiable | Engine (Tanker Shipping & Bunkering News) |
| ARA gasoil stocks, mid-September (second reading) | 12.32 million barrels, up 2% on August | 14 September 2026 | Hellenic Shipping News |
| Global observed oil inventories | Fell a further 95 million barrels in August 2026 | Reported 11 September 2026 | IEA, Oil Market Report |
| Global observed oil inventories, cumulative | Down about 507 million barrels since February 2026, roughly 2.8 million barrels a day | Reported 11 September 2026 | IEA, Oil Market Report |
The two mid-September ARA readings disagree by about 240,000 barrels; both are given here rather than averaged, because they likely reflect different weekly cuts of the same underlying Insights Global survey rather than a genuine correction. If you want to try the calculation yourself once both figures exist for the same date, the EIA's Weekly Petroleum Status Report publishes the ending-stock table and the "product supplied" table side by side, so you can divide one by the other for the US market without relying on a secondary summary.
4. Commercial stocks, the strategic reserve and floating storage are three different numbers
Commercial stocks, the kind counted in the barrel figures above, are only part of the picture; governments hold a separate layer of reserves precisely because commercial stocks move with the market and are not meant to be a safety net on their own. The EU's minimum stockholding rule, Council Directive 2009/119/EC in force since 14 September 2009, requires member states to hold the greater of 90 days of average daily net oil imports or 61 days of average daily consumption, a threshold widely shortened in coverage to "90 days." Our strategic reserve entry covers how that stock is held, released and reported in more detail; this page only tracks the day-count ratio itself.
Industry participants also use "around 90 days" as a rough historical benchmark for what counts as a comfortable level of OECD commercial stocks, though that is a long-run rule of thumb rather than a single dated statistic, and it should not be confused with the EU's legal minimum, which is a floor governments must hold, not a market comfort level. A third number that gets mixed in by mistake is oil sitting on ships rather than in a tank onshore, which our floating storage entry explains is counted separately again and does not add to any country's days-of-supply figure.
5. What this means for your own heating oil, diesel or gas bill
The same formula that describes a country's stockpile works for your own tank: take how many litres or gallons you have left, divide by how much you actually burn in a day, and you get your own personal days of supply. As a purely illustrative example, not a documented average, a household with 1,500 litres left in a 3,000-litre tank and a cold-week burn rate of around 15 litres a day would be looking at roughly 100 days of cover, a number that shrinks fast in a cold spell and stretches out in a mild one, because your daily burn rate tracks the weather, not the calendar; our heating degree days entry explains how to estimate that rate from local temperatures instead of guessing.
If your tank-days figure is short relative to your supplier's normal lead time, a small top-up rather than waiting for a full delivery slot can close the gap, a trade-off our partial fill and order size entry works through in more detail. In the UK, this same tank-days logic matters more than Ofgem's price cap, which covers electricity and gas but not heating oil, an unregulated market.
If you are in the US Northeast weighing whether to lock in an order now, our US diesel and heating oil prices page puts that timing question against the current market, and either way, put your own litres or gallons and burn rate into our fuel cost calculator instead of relying on a national barrel figure that was never about your household in the first place.
See what your own tank level and burn rate mean in days, not barrels.
Run the calculator6. What this number is not, and what people get wrong about it
A falling absolute stock in barrels does not automatically mean a falling days-of-supply figure, and the reverse is also true. If demand drops at the same time a stock draws down, the ratio between them can hold steady or even improve; if demand rises while a stock holds flat, the days figure can fall even though the barrel count in the headline stays the same. Treat the two as separate signals, not interchangeable ones, and be sceptical of any commentary that treats a barrel change alone as proof that coverage in days has moved the same way.
A second common mix-up is quoting a strategic reserve release or a floating-storage build as if it changed a country's commercial days of supply; strictly, it does not, because the metric is normally built from commercial onshore stocks, not from emergency reserves or oil still on the water. And because no verified US, OECD or Europe-wide days-of-supply number for September 2026 could be sourced for this page, treat any specific day count circulating for this period with caution unless it comes with its own dated, named source.