Module 5 · Ownership, price and failure
Reading the numbers without double-counting
Level 2 · Working knowledge 10 min read 4 questions
Production, refinery runs, product supplied, exports and inventories are five different measurements of the same chain at five different points. Adding them is the commonest error in oil commentary.
The copper material page on this site carries a warning that applies almost word for word to oil: mine production and refinery production measure different things, and adding them double-counts. Oil has the same problem in a longer chain, with more agencies publishing more series more often.
The five measuring points
| Series | Measured where | Careful of |
|---|---|---|
| Crude production | At the wellhead, after separation | Whether condensate and natural gas liquids are included; definitions differ by country and agency |
| Refinery runs / throughput | Crude entering refineries | Not the same as production: the gap is trade and stock change |
| Product supplied / apparent demand | Product leaving the primary distribution system | A proxy for consumption, not a measurement of it; stock changes downstream distort it |
| Imports and exports | At the border, crude and products counted separately | A country can be a large importer of crude and a large exporter of products at the same time |
| Inventories | Crude and each product, commercial and strategic separately | Working capacity versus nameplate; oil on water is sometimes in and sometimes out |
Definitional; applies to all vintages · Series definitions per EIA Petroleum Supply Monthly and JODI Oil methodology
Every one of these is defined slightly differently by different agencies. When two sources disagree by a few per cent, the first thing to check is the definition, not the arithmetic.
Who publishes what
- EIA
- The US statistical agency. Weekly estimates and monthly revised data for US supply, refining, stocks and trade, plus international series. The weekly is an estimate and is revised in the monthly, which is why weekly-to-weekly comparisons are noisier than they look.
- IEA
- Publishes a monthly Oil Market Report with a world supply and demand balance, from the consuming-country perspective. Its demand numbers are estimates built from many national sources.
- OPEC
- Publishes its own monthly report with its own balance and its own demand view, plus member production on both a self-reported basis and a secondary-source basis. The two bases differ, and the difference is itself informative.
- JODI
- A collaborative database of self-reported national figures. Coverage and timeliness vary by country, and gaps are not zeros.
- Price reporting agencies
- Daily assessments for grades and products, under published methodologies.
Units, and the one conversion worth memorising
- A barrel is 42 US gallons, about 159 litres.
- Production and demand are usually quoted in barrels per day; annual totals in barrels or in tonnes.
- Tonnes to barrels depends on density, so it depends on the grade. There is no single correct factor - a light crude gives more barrels per tonne than a heavy one. Any source converting between them has chosen a factor, and you should find out which.
- Crude oil and total liquids are different. The second includes natural gas liquids, condensate, biofuels and refinery processing gain, and is therefore a larger number for the same world.
A working checklist
- What exactly is being counted - crude, crude plus condensate, or total liquids?
- At which point in the chain - wellhead, refinery inlet, or product out?
- Over what period, and is it an estimate that will be revised?
- Who published it, and what is their definition?
- Is this comparable with the number I am putting it next to?
That checklist is the habit this whole course is trying to build. The physical chain has stages; every number belongs to one stage; and most bad analysis comes from quietly moving a number from the stage it was measured at to a different one.
- Production, runs, product supplied, trade and stocks measure five different points; they are not additive.
- Crude, crude plus condensate and total liquids are three different quantities.
- Agencies disagree because world demand is estimated, not measured; consistency over time beats false precision.
- Product supplied
- Product leaving the primary distribution system; a proxy for consumption.
- Refinery runs
- Crude actually processed, as distinct from crude produced.
- Total liquids
- Crude plus condensate, natural gas liquids, biofuels and processing gain.
- Copper - read the numbers correctlyon this site
- This site's data sourceson this site
- Methodologyon this site
- EIA Petroleum Supply Monthlyprimary source
- JODI Oilprimary source
Check yourself
1. A commentator adds a country's crude exports to its refined product exports and calls the total its 'oil exports'. This is:
The same molecules are counted at two stages. It is the oil version of adding mine production to refinery production.
2. Why is there no single correct factor for converting tonnes of crude to barrels?
A barrel is a volume and a tonne is a mass, so the factor is a density - and crude density varies widely by grade.
3. 'Total liquids' is a larger figure than 'crude oil' for the same country and period because it includes:
It is a broader definition, which is why contradictory 'record production' headlines can both be accurate.
4. The EIA weekly petroleum figures should be treated as:
The weekly is a timely estimate; the monthly revises it. Week-on-week comparisons carry more noise than the presentation suggests.