Friday, May 27, 2011

Global Oil Exports/Imports

Yesterday, I posted some UNCTAD shipping data which suggested that the global oil tanker fleet was expanding rapidly post 2005:

The most obvious interpretation would be that the global oil trade was expanding, which is a little counterintuitive, given that global oil production has been more-or-less on a plateau since 2005 and oil exporters have been growing their own consumption rapidly as their economies are stimulated by high oil prices.

One possibility is that shippers have been building too many ships in error, so that actual usage is not rising, even though fleet capacity is.  Commenter cimon9999 dug up this data on actual tonne-miles shipped:

This shows some increase in crude oil shipments, albeit more modest than the growth in the shipping fleet.  There is more growth in the trade in oil products.

To get a better handle on things, I then looked to the EIA who used to maintain global data on oil imports and exports (until their budget got cut).  The data run from 1986 to 2009, and look like this:

This shows an unambiguous peak in total world exports in 2005.  The average rate of decline from 2005 to 2009 was 1.8%/year.  If you believed that data, and if the 1.8% decline rate were to continue, it would take about 40 years for global oil exports to half.

However, the import data show a plateau from 2005 to 2008, and then a smaller fall (presumably due to the great recession).  Other things being equal, I'd be more inclined to believe the oil import data, since it's mostly accounted for in developed countries with reasonably functional bureaucracies, whereas the export data is coming largely from trying to infer what goes on with OPEC countries, some of whose transparency and integrity leaves a lot to be desired, in my experience, and who have an incentive to understate their production (OPEC quotas).  And the import data suggest a more optimistic picture.

Of course, those of us in the OECD will probably need to cut our usage faster than this data indicates as China, India, etc will likely continue to increase theirs.

Finally, how can the number of tonne-miles of oil shipped be increasing, even as the quantity traded decreases?  One possibility, certainly, is that one or other data set is wrong.  The other possibility is that although the amount of oil traded has declined at least somewhat, the average distance that it travels to market is increasing faster than that, so the overall volume of tonne-miles is increasing.


dr2chase said...

A few unfounded thoughts:

- if you were travelling slower (to save energy) you might want more ships.

- if you were traveling farther (to avoid pirates?) you might want more ships, and it would increase the ton-miles.

- tankers have some value if you are looking for a place to store oil for speculation.

And a brief sanity check, on cost of tanker ($120M) and capacity (200-550 KT = 400-1100 M-lb, 300 lb/bbl, gives 1.3 - 1.8 M-bbl. At $100/BBL, cost of tanker = cost of cargo, roughly). Seems to me there's a little room in the system for arbitrage.

Carrington Ward said...

There's a fair amount of commentary I've heard on shipping blogs about overbuilding in the tanker fleet, and about plummeting freight rates, which would agree with the idea that 'miles' are increasing faster than tonnes (arbitrage gets easier).

One further issue is the phenomenon of contango floating storage.

skanky said...

What about pipelines, are there any international ones that will also affect the import/export totals, without affecting shipping?

Glenn said...

I'd go with oil travelling farther. For instance, if a country goes from producing oil for it's own use as well as importing oil; as it's own productivity declines, it will import more. U.S. is an excellent example of this. Particularly as Mexico plays out and more of what the U.S. imports comes by ship rather than truck or pipeline.

cimon9999 said...

Hmm, there's a pretty large divergence in the EIA's figures (45 mb/d) and BP's Stat Rev. 2010 'Oil-Trade Movements' which shows oil movement rising from 41 mb/d in 1998 and peaking in 2007 at 55.5 mb/d, then to 54.6 mb/d in 2008 and 52.9 mb/d in 2009. This was how I worked out that average seaborne distance was actually declining, if tonne-miles have been reasonably constant since 1998.

Anyone have any insight into why the discrepancy between EIA and BP?

James said...


Excellent post. The scramble is on. Exports were about 25 Mbd in 1985; grew to 45 Mbd in 2005. Can you see an optimistic case where they reach 50, given OPEC consumption trends? Seems impossible. What's the best case then? Iraq redux, keeping them flat? What's the worse case?

Anonymous said...

I would bet on the average distance traveled by oil is increasing. Mexico's pipeline imports to the US are decreasing, Venezuela's exports are increasingly travelling the Pacific to China rather than across the Carribean to the US, etc.
Moreover, I would bet that more of the tanker fleet is now engaged in the playing of the commodities market, such as sitting off the coast, fully loaded, waiting for someone's calls or puts to come due.

Gary said...

Recall the time during the last oil price spike when oil tankers were being used as floating storage. Could it be that floating arbitrage is a big plus in a tight speculative oil market?

KLR said...

cimon9999 beat me to mentioning the BP data. They provide the previous year's numbers in each Stat Review, the issues are all downloadable back to 2001. Would provide a useful contrast to the EIA, obviously, ala the TOD piece on JODI.

The peak of oil in floating storage was ca. 225 mb, judging from graphs people made. That's just icing on the cake of total capacity. EIA used to provide real numbers for this back in the 80s/early 90s, but no longer, it seems. Morgan Downey had data on it from some arcane source.