Global Oil Supply and Demand Chart

World consumption of petroleum and other liquid fuels runs at roughly 100–105 million barrels per day, varying by a few million barrels between the seasonal low and high within any given year. The exact figure for the most recent observed month is shown at the top of this page, along with the month it refers to.

EIA Short-Term Energy Outlook Table 3e, World Petroleum and Other Liquid Fuels Consumption, retrieved through EIA API v2 (series PATC_WORLD). World supply comes from the matching production series, PAPR_WORLD. Both are monthly and are revised with each new STEO edition.

No. The series covers petroleum and other liquid fuels, which includes crude oil, natural gas plant liquids, biofuels, and refinery processing gain. That is why the total is higher than figures quoted for crude alone. EIA also notes that OECD consumption is measured as “product supplied” while non-OECD consumption is “apparent consumption”, so the two halves of the world total are not built the same way.

STEO publishes history and roughly eighteen months of projection in the same series, and does not flag the boundary in its API. This page separates them: the chart plots observed months only, and the projection is presented as text above it, clearly labelled. The boundary is located by comparing STEO’s US consumption figures against EIA’s published monthly actuals — the first month the two stop agreeing is the first month STEO is estimating. EIA itself describes the split in its printed tables as an approximate break.

Supply minus demand for the same month. If the world produced more than it consumed, the surplus went into inventory somewhere, and the bar sits above zero; if it consumed more, stocks were drawn down and the bar sits below. It is an accounting residual, not a measured inventory figure.

Less than the decimal places suggest. It is the difference between two independently estimated numbers of around 100 million barrels per day each, so it inherits the error in both. A balance of a few hundred thousand barrels a day is well within the revision range of either series. Treat large, sustained gaps as meaningful and small ones as noise.

The US holds only a portion of world commercial inventory, and stocks can build in one region while drawing in another depending on trade flows, freight economics, and refinery runs. A global balance sets the backdrop; it does not determine any single country’s weekly number. That is why the implied build or draw appears as context bars on the US crude chart rather than as a predictor of it.

Those describe how full a tank is relative to normal, and nothing on this page is a tank. Demand and supply are flows measured in barrels per day. Applying a storage benchmark to a consumption rate would produce a number with no physical meaning, so the seasonal machinery used elsewhere on StorageCurve is deliberately not applied here.

It is a considered, transparent, publicly documented projection, and it also moves substantially from month to month. In its July 2026 edition, for one documented example, EIA revised its 2026 Brent price forecast down 14% and its 2026 world production forecast up 3.5% against the edition published four weeks earlier. The direction of travel is worth knowing; the precision is softer than it looks. This page shows the projection because it is useful, and separates it from measured history because the two are not the same kind of number.

EIA publishes a new STEO once a month, typically in the first half of the month, and each edition revises both the forecast and recent history. The release date of the edition in use is shown beneath the chart.