So after the mystery of why JODI Oil regional refinery products demand data (oil products consumption) is so different from the BP Statistical Review of World Energy for 2016, I took a look at the individual country data supplied by BP and compared it to the JODI Oil numbers.
The first thing that struck me was that there are many items of data that are very similar between the BP and JODI Oil data; and yet there are also a good number that are significantly different – and the vast majority of these show BP reporting much higher oil consumption than JODI. This means that the definitions that BP and JODI are using for oil products consumption must correlate in many cases, when countries make their reports. But it also means that there are some understandings of oil consumption that BP has that do not have cognates in the JODI Oil reports.
The second thing that struck me was that each region in BP apart from North America is showing a total much higher than JODI Oil. Only some of the countries are specifically named in the BP report, and other countries are lumped into the bucket of “Other” within each region. Each “Other” figure is much higher in the BP report than in the JODI Oil data. Part of the reason is clearly going to be because some countries have not been reporting to JODI Oil, or not reporting reliably. For example, for South and Central America, JODI Oil data for Bermuda, Cuba, El Salvador, Haiti and Suriname are all zeroes; and JODI Oil data for Bolivia has zeroes for NOV2015 and DEC2015 (other months average at 63 KBD). But these could all be expected to be low oil products producers; so it is unclear to me where BP thinks consumption is occurring outside of the individually-named countries.
The “Other Africa” line is much higher in BP than in JODI, which looks dubious. I have not looked at this closely, but this might relate to countries such as Nigeria who produce and also consume a lot of oil.
The most significant differences : countries where no JODI Oil data is available : Turkmenistan, Uzbekistan, Israel, Bangladesh, Pakistan; and also countries with medium-to-high BP oil consumption data compared to JODI : Brazil, Venezuela, Belarus, Kazakhstan, Russian Federation, all the named Middle East countries, South Africa, China, India, Indonesia, Malaysia, Singapore, Thailand, Vietnam.
It could be that in some cases the BP data is for all oil consumption – from national refineries and imports; whereas the JODI Oil data is for consumption from a nation’s own refinery. I would need to check this in more detail, but at first glance, the BP oil consumption data for the Middle East is much more divergent from the JODI Oil data than for other regions, and this does not make sense. I know that refinery product self-consumption is increasing in Middle East countries that are in strong economic development, but not all Middle East countries are experiencing increasing national demand, and I cannot imagine that oil products imports are so high in this region as to explain these differences between BP and JODI Oil data.
Another thing to note is that Commonwealth of Independent States (CIS) (formerly known as the “Former Soviet Union”) data divergence accounts for most of the data divergence in the “Europe & Eurasia” region; and that BP oil consumption data for the Russian Federation (which forms a part of CIS) is much higher than the data given to JODI.
I now have too many questions about how and from whom all this data is sourced, how categories of liquid hydrocarbons are delineated, and doubts about how anybody could check the reliability of any of this data. Without more information, I cannot analyse this data further; but maybe looking at oil consumption is not that illuminating. There appears to be a small and steady increase in annual oil demand and consumption over the recent period – this is indicated by both BP and JODI Oil data. The real issues for my analysis are whether oil production is capable of sustainably satisfying this demand-with-small-annual-increases, so my next step is to move to look at liquid hydrocarbons production data.