This is my first attempt to reconcile the 2015 global oil production data from five different publicly available sources : JODI Oil, BP, OPEC, EIA and IEA, and truth be told, it’s ugly.
I can feel I’m going to need to redo every step, just in case I made an error in assumption or copying figures into my spreadsheet(s).
I’m also going to need to contact each of the agencies for one reason or another, in particular to request a country-by-country full breakdown of the data, as it is impossible in some cases to compare the regional country groupings used by each agency.
In order to do this comparison, it has been necessary to read the “fine print” in the data reports and database information from the agencies, to try to understand how each of them treats each territory it holds data for, and which geographical region it assigns to which data for each country. A couple of notes here should show how complicated it can get : for example, BP considers Mexico to be a part of “OECD Americas” and “North America”, but OPEC considers it to be in “OECD Americas” and “Latin America”; the EIA consider Estonia to be a part of “Eurasia” in recent data downloads, whereas BP considers it a part of “OECD Europe” in the Statistical Review of World Energy 2016; and OPEC includes data from Indonesia in its total of OPEC oil production for 2015 in the Annual Statistical Bulletin 2016, but Indonesia only rejoined OPEC on 1st January 2016.
1. JODI Oil Data
I downloaded this data in late May 2016, and ran it through a C programme to group the country data roughly according to the BP schema.
2. Missing JODI Data
Where country data was missing in JODI, I filled in the gaps by pulling out the figures from the EIA Crude Oil (including Lease Condensate) data. I chose this data set because a comparison of figures between JODI Oil and EIA for the United States showed they were close. This I call “Adjusted JODI” data.
3. Regrouped Adjusted JODI Data
I re-grouped the Adjusted JODI data to match the regional groupings of the other data sets – essentially pulling “OECD Asia Pacific” and “Other Asia” data into the same group.
4. OPEC Annual Statistical Bulletin
I took the data for OPEC oil production from the OPEC ASB 2016 Table 3.5 and for the rest of the world from OPEC ASB 2016 Table 3.7. I then compared OPEC and JODI Oil data by subtracting the JODI data from the OPEC ASB data. Since some of the countries were not specifically named, and belonged to different regions in the JODI analysis, the results are not completely accurate. It was not possible to split “Eastern Europe and Eurasia” into “Europe” and “Eurasia” countries.
5. Adjusting OPEC ASB data for OPEC countries
The OPEC data for OPEC countries does not report Lease or Field Condensates in the main crude oil figures – these are lumped in with NGL figures, which also include NCF – non-conventional fossil fuels. The OPEC data for non-OPEC countries appears to include NCF in the main crude oil figures. The JODI Oil data do not appear to include NCF. So for the regions where there were significant NCF showing in the EIA data, I added these on to the JODI figures to permit a clearer comparison to the OPEC data.
6. IEA Oil Market Report (OMR)
I took the 2015 data from the International Energy Agency (IEA) OMR of 13 July 2016 and compared them to the Adjusted JODI data. The difference for the OPEC figure seemed very large, and this appeared to be because NCFs were included for the OPEC data, but not in the other figures. So I subtracted the OPEC NGLs figure from the IEA OPEC total, and instead added in the NGLs figure from the EIA data for the comparison with JODI.
7. EIA Data
I compared the Crude Oil plus Natural Gas Processing Liquids (NGPLs) data from EIA with JODI.
8. BP Data
I compared the BP Statistical Review of World Energy 2016 page 8 for oil production in thousands of barrels per day with the JODI data. I needed to move some of the countries between regions for the comparison, but this was not possible as they were not explicitly mentioned in the BP data – splitting “Europe and Eurasia” into “OECD Europe”, “Eurasia” (Former Soviet Union or FSU) and “Other Europe”.
My main conclusion so far is that anybody basing analysis on any of these data sets should be very wary. Some of the numbers look suspect. Also, the total production of hydrocarbons may be larger than previously, but it’s an apples and oranges problem : NGLs are not the same as crude oil, and cannot give the same amount of refined oil products.
In my next post, I’m going to explain all the acronyms I haven’t explained this time, and delve further into regional geography.