Friday, March 23, 2012

Part 3: Inter-Regional Trade Movements of Petroleum in North America

In Part 1, I introduced my  abbreviations, data bases and analysis methods, and in Part 2, I presented the global trends. 

Figures 1-6 show the data somewhat analogous to the global trends presented in Part 2, while Figures 7-10 show the inter-regional petroleum export trends to and from North America (NA) which is unique to the information provided in BP reviews for the years 2000-2010.

North American Gross and Inter-regional Export and Import Trends

As I explained in Parts 1 and 2, the EIA, just provides import and export data for individual countries which I can sum up to correspond to my nine global regions.  As such, for regions with multiple countries (i.e., all regions except China and Japan) these sums will include intra-regional and inter-regional import and export data, which I call “gross” imports or exports.

In contrast, the data in the BP review allows me to derive inter-regional total petroleum imports and exports for each region.  I also derived estimates of inter-regional crude and product imports and exports, by assuming that my correction factors for the totals can be equally and proportionally applied to the subtotals for crude and products.

Figure 1 presents NA’s total petroleum consumption (i.e., both domestic and imported), the total imports, and, the gross and inter-regional imports into crude and petroleum products, from the data reported by the EIA and BP. 

As you can see in the figure, the BP and EIA total consumption numbers (red circles and squares at the top) are in close agreement with each other and illustrate the long-term increase in petroluem consumption in NA since 1983, and more recently, a 4-year plateau at 9.1-9.2 bby from 2004-2007, and a down-turn thereafter.

Total gross imports (open red circles) peaked in 2005, at about 5.15 bby, or 56% of total consumption, while total inter-regional imports (open red squares) peaked in 2007 at about 4 bby, or 43% of total consumption.

The gross crude imports (open brown circles) peaked at 4 bby in 2005 although there is a plateau at about this value from 2004 to 2007.  The inter-regional crude imports (open brown squares) also show a plateau from 2004 to 2007 at about 2.9 bby and peak in 2006.

The gross (open green circles) and inter-regional petroleum product imports (open green squares) are very similar to each other, implying that the intra-regional imports must be very small at least in years of data overlap (2000 to 2009), which I will discuss later (Figure 5). There are some indications of a shallow peak in 2005 and 2007 of gross and inter-regional imports, respectively.

Figure 2 presents NA’s total petroleum production (i.e. domestic production, blue circles and square for EIA and BP data, respectively), crude oil production (open purple circle), and the gross and inter-regional total petroleum, crude and petroleum products exports. 

Figure 2 illustrates that NA's total petroleum production has been in a fairly narrow range from 5 to 6 bby since 1980, although there is a trend for a decline since a peak in 1985 (5.9 bby, EIA; 5.6 bby BP), until a recent up-tick in the last two years.  Crude production continues a more pronounced decline from its peak of 4.8 bby in 1985, implying, I think, that the recent up-tick is mainly due to other products like NGLs or biofuels.  Or, perhaps this is a reflection of shale oil production domestically, which I would think that this would be counted as part of crude oil production. Indeed the last two EIA reporting years for domestic crude production (2009-2010) does should a small up-tick.

Despite these overall rather flat or declining domestic production trends, NA's gross and inter-regional total exports (red circles and squares) are in long term upward trends.  Gross crude exports (brown circles) peaked in 2006, while inter-regional crude exports have been slightly rising over the last ten years. This might imply a recent down trend in intra-regional crude exports, but I will talk about this later (Figure 6).  Gross and inter-regional petroleum product exports are in relatively shallow upward trends.

In part 2 (Figures 2 and 3), I showed that the total global import and export pools of petroleum showed signs a plateauing and peaking around 2006-07, and so it is interesting for me to look at regional trends relative to these diminishing global import and export pools.

Figure 3 shows NA's imports relative to the respective total gross and inter-regional global import pools. Note that percentages for the gross import values only runs to 2008, because that is last year when the EIA reported global values, and that is the number I need to calculate a global import pool with which, in turn, I can use to calculate a percentage. Also, as you look at Figure 3, keep in mind that these are NA’s imports as a percentage of the gross global import pool, for the EIA derived data, and, the global inter-regional import pool for the BP derived data, respectively.

The gross global imports are bigger than the global inter-regional imports.  For example, from Figure 2 in Part 2, in 2008, the total gross global pool was 22.6 bby, while the total global inter-regional import pool was 17.1 bby.  This helps explain why, in Figure 3 above, NA’s relative share of the global inter-regional imports is higher than NA’s relative share of the gross global import pool, even though the absolute value is lower.

NA's relative shares of imports generally show more prominent signs of peaking than the absolute values in Figure 1.

NA’s gross total petroleum imports as a share of the total gross petroleum import pool (red circles) peaked in 2001 at 22.5%, and then bumped along a plateau (with a  prominent dip in 2002, post-911) for 4 more years before dropping down to 20.3 % by 2008, the last reporting year that I can calculate a percentage.  NA’s inter-regional total petroleum imports as a share of the total inter-regional petroleum import pool (red squares) almost peaked in 2001 at 24.1%, but then rose to a value slightly higher than this by 2005 to 24.2%, and then dropped down to 19.2% by 2010.

NA’s gross crude oil imports as a share of the total petroleum import pool (brown circles) peaked in 2001 at 18.4% and then gradually decline to 17.1% by 2008.  NA’s inter-regional crude oil imports as a share of the inter-regional petroleum import pool (brown squares) was actually at its highest value in the reporting period in 2000 at 18.7%, and has declined since then to 14.2% by 2010.

NA’s gross petroleum product imports as a share of the total petroleum import pool (green circles) actually peaked way back in 1988 at 5.8%, although it rose again to local peak in 2005 of 4.9%. At least from 2000 to 2010, NA’s inter-regional petroleum product imports as a share of the inter-regional petroleum import pool (green squares) plateaued from 2005 to 2007 with a peak of 6.6% before declining to 5% in 2010. 

Figure 4 shows NA's exports relative to the respective total gross and inter-regional global export pools, which, of course, is about the same as the global import pool.  Analogous to Figure 3, percentages for the gross export values only run to 2008.  Also analogous to the imports trends shown in Figure 3, these relative export amounts are expressed as a percentage of the total gross export pool and the total inter-regional export pool, and, these amounts are different: e.g., 23 bby of gross exports in 2008 versus 17.1 bby inter-regional exports in 2008.  This time however, NA’s gross relative exports are greater that the inter-regional exports.

I kept the same vertical scale as used in Figure 3 to make the point clear that NA’s petroleum exports are far less, about 3 times less, than its imports.  NA is a heavy net importer, although that trend is changing as we will see below.

Since 1986, NA’s gross total petroleum exports as a share of the total gross petroleum export pool (blue circles) has been in a range (e.g., 8.6% in 1991 to 7.3% in 2008) of total global exports, with a slight downward trend. On the other hand, since 2000, NA’s share of inter-regional exports have been in an upward trend (blue squares) increasing from 2.5% in 2000 to 4.1% by 2010.

NA’s share of gross crude exports are in a slight downward trend (5.7% to 4.6% from 1986 to 2008) and inter-regional crude exports slightly trending upwards (1.8% to 2.2% from 2000 to 2010).  

NA’s gross petroleum product exports have been in a range of 2.4 to 2.6% since 1986, while inter-regional product exports are trending upwards from 0.8 % in 2000 to 1.9% in 2000.


The difference between the gross and inter-regional imports or exports, shown in Figures 1 and 2, respectively, should correspond to intra-regional movements of petroleum.

For instance, Figure 5 shows the differences between the gross and inter-regional total, crude and product imports, which should equal the intra-regional values for these three quantities.  

The total (red triangles) and crude (brown triangles) intra-regional imports are both about the same at 1-1.2 bby.  As I noted above, this implies that the intra-regional imports must be very small, which is reflected in Figure 5 as product intra-regional products (green triangles) being close to zero.  The negative values for implying that the intra-regional imports from 2006-09 don't really make sense, because the gross product imports must always be greater then the inter-regional imports. 

Figure 6 shows the differences between the gross and inter-regional total, crude and product exports, which should equal the intra-regional export values for these three quantities. 

The total exports (red triangles) are about 1.2 bby with some large fluctuations in the last three years.  But unlike the imports, crude exports (brown triangles0 are substantially lower than the total and in somewhat of a downtrend.  Also, in contrast to the imports, the trend is for intra-regional  product exports to be increasing with large spike upwards in 2009, reflecting the spike in gross production export shown in Figure 2.

In principle the intra-regional import and exports should be identical, and, these differences in between Figures 4 and 5 illustrate the limitations in my procedure for deriving intra-regional changes.  As I explained in part 1 and 2 I am comparing the difference between two different data bases (EIA versus BP) and I am also deriving the crude and product exports from the BP data, by assuming that my factor for correcting for inter-regional imports and exports would be proportionally distributed between crude oil and petroleum products. The discrepancy between NA’s product import and product exports suggests that these assumptions are probably breaking down here.  In particular, there is probably a proportionally larger amount of intra-regional product imports than this procedure has not accounted for, and, I know of no way to account for.  Additionally, as I pointed out in Part 2, and it is also illustrated in Figure2, above there seem to be a systematic difference between the EIA's and BP's estimates of total production, and perhaps these systematic differences carry over to the import and export data. 

A final cross check of the calculation of intra-regional exports and imports
As a final check of the intra-regional import and export estimates presented in Figures 5 and 6, I used the trade movement data provided for USA, Canada and Mexico in the BP review. 

These data provide a unique means to cross-check the total intra-regional petroleum export and import data.  Because NA essentially comprises only three countries: the USA, Canada and Mexico, and the trade movements of petroleum are explicitly reported in the yearly BP reviews, I can explicitly calculate the total intra-regional imports and exports for NA.  This is unique as compared to any other my other defined multi-country regions (i.e., SA, EU, APr, ME, FS), because the individual movements between countries within those regions are not reported by BP. 

An example of how to calculate NA's total intra-regional exports and imports from the BP trade movement data, both equal to 4452 kbd or 1.62 bby for 2010, was presented in Part 1.  Analogous calculations were made for 2000-2009 and these data are presented in Figures 5 and 6 (orange triangles). 

From 2000 to 2005 there reasonable agreement between the difference (EIA gross minus BP inter-regional), and, intra-regional imports and export calculated directly from the trade movement data presented for US, Canada and Mexico.  But, thereafter, these two estimates of intra-regional imports and exports diverge.  Whereas the difference (EIA gross minus BP inter-regional) dips down the intra-regional imports, the exports calculated directly from the trade movement data stays fairly steady.

I give more creditably to the intra-regional values calculated directly from the trade movement data, because the trade movements between the three major countries in NA are explicitly reported in the yearly BP reviews, whereas the difference calculation is relying on two separate data sets that seem to have inherent differences.

These discrepancies are large enough to persuade me to reduce my future efforts at calculating intra-regional import and export trends for the other regions, using the difference between EIA gross minus BP inter-regional import/export data.  I will continue to show both gross and inter-regional imports and exports, but I will reduce the amount of time I spend on analyzing the differences between these two numbers.

Trade movements of total petroleum between North America and other regions

Figure 7 show the specific amounts of petroleum imported to NA FROM each of the eight other regions.  For reference, Figure 7 also shows NA's total (i.e., crude plus product) inter-regional petroleum imports from all 8 of the other regions (black "Xs"). This is the same as the data presented in Figure 1 (red squares).

As noted before in context of Figure 1, NA's total inter-regional imports plateaued from 2005 to 2007 and declined thereafter.  This trend is mainly reflected by the changes in imports from five regions: EU, AF, ME, SA and FS.  Absolute imports from EU, SA and ME are trending down while imports from AF and FS are trending up.  Imports from APr, JP and CH are very small (about 0.1 bby or less from all three regions for any of the years from 2000 to 2010).

In the year 2000, SA was NA’s top import source at 0.99 bby with ME very close behind at 0.96.  AF and EU were third and fourth, respectively, at 0.65 and 0.53 bby.  FS was a very distant sixth account for only 0.02 bby—the remaining Asia Pacific countries (APr)  were actually more important importers than FS at 0.08 bby.

By the year 2010 AF was the top importer at 0.92 with SA still in second at a slightly lower 0.85 bby.  The ME essentially swapped place with AF now at third at 0.67 bby.  EU is still the fourth importer at 0.37 bby, but if these trends continue, it will soon be eclipsed by FS whose imports to NA have increased over the last decade by over an order-of-magnitude since 2000 to 0.29 bby.

Figure 8 shows NA show the specific amounts of petroleum export from NA TO each of the eight other regions and again for reference, I show NA’s total exports (black “Xs” corresponding to the blue squares in Figure 2).  Note the different vertical scale compared to Figure 7.

As noted previously in the context of Figure 2, NA’s total exports are in a strong up-trend through the entire decade.  Where are these exports going to?   Well in the year 2000 NA exports went to EU, SA, APr and JP in that order.  By 2010, SA was the top destination for NA’s exports with EU and APr in second and third.  JP is still the distant fourth highest export destination, but it is about to be replaced by CH or AF which have risen from negligible imports from NA five years ago to nearly the same a JP by 2010.  Not surprisingly NA doesn’t export to ME or FS.

Figures 9 and 10 present the same data as shown in Figures 7 and 8, respectively, but expressing NA's imports or exports, to or from each of the eight regions, as percentages of the total global inter-regional import/export pool (of course, on a global scale inter-regional imports and exports are the same) .  For reference, I also show NA imports and exports as percentages of global import/export pool (“Xs” right vertical axis; note the different scale). 

Additionally, I have taken all of these data and made linear extrapolations of the 2000 to 2010 trends (linear regression analysis) out to 2021.  The r2 values for the linear regression analysis result shown in Figure 9 are pretty large for all of major import source regions:  0.94 for FS, 0.90 for ME, 0.85 for SA and 0.83 for EU.  For other regions AF, APr, CH, r2 ranges fro 0.42 to 0.70, and for JP whose slope is nearly zero, r2 was 0.02.  For NA's total relative imports r2 was 0.6.  The r2 values for the linear regression analysis result shown in Figure 10 and not quite as large for the major export regions:  0.78 for NA, 0.81 for SA, 0.19 for EU, 0.78 for AF, 0.20 for APr, 0.45 for CH, and essentially 0 for JP, ME and FS.

The percentages and linear regression lines in Figures 9 and 10, I think, nicely shows the over-all trends for decreasing inter-regional imports and increasing inter-regional over the last decade. 

As illustrated in Figure 9, in 2000, NA was importing 24% of the total global import/export pool.  And, 7.2%, 6.9% and 4.7% of that 25% was coming from SA, ME and AF, respectively, corresponding to 78% of NA's total imports.  EU was supplying 3.8% of that 25%, corresponding to 15% of NA's total imports.  

By 2010, NA was importing 19% of the total global import/export pool, and 5.6%, 5.1% and 4.0% of that 19% was coming from AF, SA and ME, respectively, corresponding to 77% of NA's total imports.  EU was now supplying 2.2% of that 19%, or 11% of NA's total imports. 

If the linear extrapolations of the last decade's trends are correct, then by 2021, NA will be importing slightly less than 16% (15.7%) of the total global import/export pool.  Of that 16%, 7.8%, 3.2% and 3.1% will be coming from AF, FS and SA, respectively, corresponding to 90% of NA's total imports.  Notably, AF is predicted to provide almost half of NA total imports.  EU is only providing 0.4% of that 15.7%, or 2.5% of NA's total imports.  And imports from ME are almost as low as the EU, providing 0.67 of the 15.7%, or only 4.2% of NA's imports. 

As illustrated in Figure 10, in 2000 NA was exporting 2.5% of the total global inter-regional import/export pool.  Of this 2.5%, 1.1% and 0.9% of this went to SA and EU respectively corresponding to 80% of NA inter-regional exports. 

By 2010, NA's exports amounted to 4% of the total global inter-regional import/export pool, and again most of this went to SA and EU.  Of the 4%, 1.8% and  1.1% went SA and EU respectively corresponding to 72% of NA's exports.  APr received 0.4% or 10% of NA exports. 

Assuming the linear trend depicted in Figure 10 hold, by 2021 NA's exports will amount to 5.2% of total global inter-regional import/export pool, with SA EU and APr receiving 2.4, 1.4 and 0.6 of this 4%, or about 85% of NA's exports.  Curiously, around 2106, NA will export more petroleum than it imports from EU, and therefore, NA will become a net exporter to EU. 

Figures 11 and 12 show the relative changes in NA's import sources and export destinations, respectively, as a percentage of NA's exports or imports in the years 2000, 2010 and 2021 (predicted).


Summary and Conclusions
North America is a large importer of petroleum, but there are signs that its absolute imports peaked around 2005-2007, and imports are in decline.   In 2000 and 2010 North America imported about 24% and 19% of the total global import/export supply pool, respectively; by 2021, this is predicted to be down to about 16%.  Over the last decade, North America's chief suppliers of petroleum have changed in order of importance, from South America, the Middle East and Africa in 2000, to Africa  South America and the Middle East in 2010.  If the last decade's trends continue, then by 2021, North America's main suppliers will be Africa, the Former Soviet Union countries and South America, in that order. 

Although it is a large net importer of petroleum, North America does export petroleum, and the amount of these exports have been increasing.  Whereas it exported about 0.35 bby in 2000, by 2010 this was up to 0.7 bby in 2010.  The bulk of the increase has been due to increase product exports.  These amounts correspond to 2.5% and %4, of the total global import/export supply pool, respectively.  And the trend continues North America would be responsible for about 5% of the global supply pool by 2021. The main destinations of North America's petroleum exports have been South American and Europe and it looks like this trend will continue. 

Why have North America's imports gone down and exports gone up?  Well, declining imports are probably at least impart due to the plateau and then decrease in the total global import/export supply pool since 2006 (see Figure 3 in Part 2).  Additionally, the decline in the economy, increased unemployment, and the aging population are also likely factors explaining decreased consumption.  Finally, increased consumption in other areas like China and other Asia Pacific countries is also competing for that diminishing import/export pool, driving up the cost of petroleum.

The lower domestic consumption and increased prices probably also explains why North America's petroleum exports have increased and are likely to increase in the future.  If petroleum is truly a fungible commodity, then it will go to the regions that can afford to pay the highest price for it.  Increasingly, that region is not North America. 

3-24-12: Although it is a wonderful tool, without which probably none of the material in this blog would have been published, BLOGGER does have its challenges.  One of those challenges is that what looks like a perfectly fine figure on my screen, doesn’t always look so good once published on BLOGGER.  I went back and bolded and increases the size of all the legends and axis labels of all the figures so that they don’t look so washed out.

3-25-12 I added figures 11 and 12 to better show the trends being discussed.

3-31-12 correcting a typo in figure 1
----------------------
I will stay in the Western Hemisphere, and next time examine South America's Export and Import Trends. 

Saturday, March 10, 2012

Inter-Regional Trade Movements of Petroleum: Part 2 Global Trends in Exports and Imports

With the house-keeping chores of Part 1 out of the way, let’s look at the global trends.  In subsequent parts of this series, I will described the petroleum export and import trends for each of the nine regions, which I have already analyzed and summed up in order to arrive at the world-wide trends described here. 

These data suggest that gross and inter-regional global imports and exports of petroleum and crude oil peaked five years ago in about 2007, or earlier in some cases.

Global Trends in Total Petroleum and Crude Oil Production and Petroleum Consumption
I will start with the world-wide total petroleum and crude production statistics, as reported by the BP review (2011) and the EIA (Figure 1).


BP and the EIA report production in units of thousands of barrels per day, but I find it more easier, or maybe I am just used to, expressing production in units of billions of barrels per year (bby) which is how I report all of the production, consumption, import and export data here and future parts of this series.

The EIA’s International Energy Statistics division reports production as “total oil supply,” from 1980 to 2010.  The EIA defines the total oil supply as including, “the production of crude oil, natural gas plant liquids, and other liquids, and refinery processing gain” (See Table Notes).  Crude Oil data for Canada include oil processed from Alberta oil sands.

The BP review provides production statistics from 1965 to 2010, but I only present the data from 1980 to 2010 here. BP defines production as: “crude oil, shale oil, oil sands and NGLs (the liquid content of natural gas where this is recovered separately)” and it “Excludes liquid fuels from other sources such as biomass and coal derivatives.”  (BP review 2011 footnote, p.8).  However p.39 of that same review says “Consumption of fuel ethanol and biodiesel is included in oil consumption.”  Therefore, to be consistent with BP’s consumption numbers, I added “biofuels” production (newly added in the 2011 review but backdated to 2000) to BP's oil production statistics, to produce a "total production" statistic, a sum that is probably closer to the “total oil supply” reported by the EIA. 

Turning to Figure 1, EIA’s total supply, consumption and BP’s consumption are all in good agreement with each other.  The BP total production is slightly lower (about 1 bby or 3% in 2010) as compared to the other three statistics, and, this is despite adding in the biofuels.  The discrepancy bugs me (as it has in the past), but I can’t see a calculation error on my part. 

Regardless of which production statistics we are looking at, however, the fall-off in the rate of production increase (i.e., deceleration) since about 2005 is notable.  For instance, using the EIA production data, for the 5-year spans from 1986 to 1990, 1991 to 1995, 1996 to 2000, and, 2001 to 2005, production increased by a total in a range of 1.5 to 2.5 bby per each of these 5 year spans.  The corresponding BP production data indicates similar 5-year increases in the range of 1 to 2.5 bby.  For the last 5-years, however, from 2006 to 2010, the increase is only 0.8 (EIA data) or 0.5 bby (BP data). 

For crude production, the deceleration is even more prominent: the 5-year changes from 1986 to 2005 range from 1 to 2 bby, but in the last 5-year period, the increase was only 0.2 bby per 5-year period (EIA data only available); essentially then, global crude oil production has been in a plateau since 2005.

The greater slow down in crude oil production as compared to total petroleum production means that those other liquid fuels, “natural gas plant liquids, and other liquids, and refinery processing gain” and biofuels, are likely making up a larger proportion of production, and crude oil, a smaller proportion, than in the past.  For instance, using the EIA statistics, crude oil accounted for 91% of total petroleum production in 1985.  By 2010, crude oil accounted for 85% of total petroleum production.  Not a big change but the trend is there.

Global Trends in Total Petroleum and Crude Oil Exports and Imports
While the BP review only reports total production statistics, and not crude oil production, as I discussed in Part 1, it does report total and crude exports and imports.  And, as also discussed in Part 1, these numbers can be corrected to provide an estimate of inter-regional export and import trade movements for each of the nine regions of interest to me.   Here I only report the world-wide numbers, which is the sum of these nine regions.  The individual regions, I will report on later.

The EIA also "used to" report statistics on both total petroleum and crude oil exports and imports, but as discussed in Part 1, these are “gross” exports and imports, which means that they do not exclude intra-regional trade movements. 

Figures 2 and 3 present the import and export data, respectively.  For reference, I retained the solid lines corresponding to the petroleum and crude oil production statistics, previously shown in Figure 1. 


I will not spend any time comparing Figure 2 to Figure 3, other than to note that the BP inter-regional total, crude, and product exports, are all identical to their import counter-parts.  The EIA gross total petroleum and crude oil import and export data are very similar, but not quite identical. 

For the rest of this post, I will just focus on the export data represented in Figure 3.

The most noteworthy thing about the trends in Figure 3 is that the gross total petroleum exports and inter-regional total petroleum exports peaked out in 2006 and 2007, respectively.  Gross crude oil exports peaked in 2005, and inter-regional crude oil exports peaked in 2007. 

Gross product exports, calculated as the difference between the EIA’s total petroleum exports and crude oil exports, peaked in 2006, while the inter-regional exports have not peaked, although there are signs of a plateau since 2007 (e.g., 2010 product exports of 4.03 bby is only 0.01 bby greater than the 2007 product exports).

Figure 3 presented the export data as absolute values in units of bby.  Figures 4, 5 and 6, respectively, present the gross and inter-regional exports of total petroleum, crude oil and products as percentages of total production.  The EIA-derived gross exports were calculated as a percentage of the EIA-reported total oil supply as shown in Figure 1, and the BP-derived inter-regional exports were calculated as a percentage of total production determined from the BP data, as shown in Figure 1. 

The peaks in relative gross and inter-regional total export petroleum in 2006 and 2007 are quite prominent in Figure 4, as are the analogous peaks in crude oil exports, in 2004 and 2007, respectively in Figure 5. 


I find it interesting that gross exports, that is, intra-region plus inter-region exports, peaked out at such a high percentage of total production; about 77% in 2007!  Not only is petroleum vital to supporting global trade, petroleum itself is a highly traded commodity, it seems.  Peaking in 2004, the gross crude oil exports corresponded to 52% total production, an incredible high proportion of a raw material to be exported from one country to another country or region. Similarly, the inter-regional trade movement of petroleum, that is, trade between more distant countries, peaks at 57% of total petroleum production and 44% of crude oil relative to total petroleum production. 

The long-term upward trend in petroleum product exports, both gross and inter-regional, is evident in Figure 6.  Although the gross product exports peaked in 2006-07 at 26% of total petroleum production, the inter-regional exports continues to increase, and most recently equaled 13% of total production in 2010. 

As I mentioned at the end of Part I, the differences between the EIA’s gross exports, and BP’s inter-regional exports should equal intra-regional exports.  These two data set overlap for several years from 2000 to 2008/9 and so we can derive some insights into intra-regional exports for these years, at least. 

The differences between the gross and inter-regional total petroleum exports (Figure 4), crude oil exports (Figure 5) and petroleum product exports (Figure 6) are all summarized in Figure 7.

Total intra-regional exports (red circles and line) peaked in 2004, as did crude oil intra-regional exports.  Petroleum product intra-regional exports peaked in 2006. 

Summary and Discussion
The results of this analysis suggest that peaks in the exports of total petroleum and crude oil occurred in about 2007 or earlier (Figure 3).  Gross petroleum and intra-regional product exports peaked in 2006 (Figure 7), although inter-regional product exports has not yet peaked (Figure 3).

Expressed as a percentage of production for each year (Figures 4-5), the sharpness of the peaks in total petroleum production and crude oil production are generally more prominent than the trend in absolute values (Figure 3).  This likely reflects the observation that, while rates of production have leveled off over the past 5 years, rate of exports have actually declined over this period, at least for total petroleum and crude oil production.  The "other" petroleum product production and exports is lagging this trend, and so as a percentage of production, the peak is not as evident.

I find it interesting that the trend is for intra-regional exports to have peaked earlier (Figure 7), and to have peaked more sharply, than for the peaks in inter-regional exports (Figure 3, squares).  Gross exports (circles in Figure 3), as the sum of the intra- and inter-, of course, is somewhere in between.

I find this interesting because it run counter to one of the assumptions I made in my earlier series, Estimating the End of Global Petroleum Exports, done nearly a year ago.  In that series, my "Regionalism" assumption was that all intra-regional petroleum consumption is met by intra-regional petroleum consumption and then the excess is exported or the deficiencies are imported.  I knew at the time that this was an approximation of the reality of exports and imports, but a reasonable one.  Reasonable because of the added cost of exporting oil or products outside of one’s geographical region, and, the benefit promoting political and social stability with one’s closer intraregional neighbors.  If that was what hade been happening recently, however, then intra-regional exports would have been maintained in favor of inter-regional exports.  That is does not appear to be happening, perhaps suggests a way to improve my previous estimate of the end of global petroleum exports and its implications.

Perhaps the earlier peaking of intra-regional exports reflects the ability of the exporting countries, or regions, to command a higher price, and hence profit, through inter-regional exports, as compared to intra-regional exports.  I certainly can imagine how this could be the case for exporters located in South America (SA) and Africa (AF), and maybe even the Middle East (ME) and Former Soviet Union (FS).  For instance, net exporters like Brazil and Venezuela in SA, and, Nigeria and Angola in AF, can probably fetch a higher price for their oil by selling to North America (NA), Europe (EU), Japan, China, and the rest of Asia (APr) than selling to other net importing countries in SA and AF, respectively.  Likewise, net exporters like Saudi Arabia can maybe sell their oil at higher price out outside of their geographical region than to net intra-region importers in the ME like Israel, Jordan or Lebanon.

My hope is that each of these regions will have a discernable trends in intra- and inter-regional exports, and that I could use these trends to more accurately predict how petroleum exports will change, and therefore how total petroleum consumption will change, going forward.

Another assumption that I had made in my earlier series from a year ago (Part 5 Predicting regional petroleum consumption in a post-export world) was that going forward, the proportions of petroleum imports to AP, EU and NA would all stay the same as they have been over the part 5 years.  This simplifying assumption has some basis, in that NA’s and EU’s import trends were fairly flat.  AP’s import trend was increasing but, my assumption was that this would have to flatten out too, if the exports from bet exporter regions, like the ME, AF, FS and SA, were to all flatten out, or go down, as I expected. 

My hope is to explore each region for trends in changes in the inter-regional trade movements of petroleum and project these trends forward to enable a better prediction of how each region’s imports and exports will change in the future. 

-------------------------------
next up, Inter-Regional Trade Movements of Petroleum in North American

Sunday, March 4, 2012

Inter-Regional Trade Movements of Petroleum Part 1: Introduction

Global flows of petroleum are vital to supporting a global economy, as petroleum is the key transport fuel that allows the movement of people, raw material and finished goods between distant regions of the world.  If the inter-regional flows of petroleum slow down, then so to must the global trade of goods.

This is the start of a new series of posts to summarize my analysis of changing trends in inter-regional movements of petroleum, including crude oil and petroleum products.  In this post, I summarize my data sources and methods of analysis.

Data Sources
1) The BP review
Every year BP (aka British Petroleum, aka Beyond Petroleum) for at least the last decade,  publishes on-line its “Statistical Review of World Energy” (hereinafter, BP review), a 40 to 50 page extensive summary of all the major energy sources and their consumption for all of the regions of the world and selected individual countries.   Included in the Petroleum section of this report are two tables entitled, “Inter-Area Movements” “Imports and Exports”  reproduced below from the 2011 review:


Accompanying these tables is a rather complicated chart summarizing some of the Inter-Area Movements. 

There is just way too much going in this chart to make much sense of it.  After staring at it for a while, you can get the sense that there are flows from Africa and the Middle East to North America and Asia Pacific regions, and extensive flows between North and South America and within the Asia Pacific region and a large region called Europe & Eurasia, which includes European and the Former Soviet Union countries.  But, after awhile my eyes glaze-over, I promise myself to go back and make more sense of this chart and these tables, but never do. 

Until now, that is.

I am interested in analyzing the details of inter-regional changes in petroleum imports and exports over time.  That is, how much petroleum is flowing between the major regions of the world and what is the time-trend of these interregional flows? 

The BP review does provide a third table entitled, “Trade Movements” which shows time dependent imports and exports for selected regions.  However, this table only shows imports for the USA Europe and Japan, with the rest of the world all lumped into one category.  A more extensive list of exporting regions or countries are also presented in this table, but, without knowing the imports into these regions or countries, one can’t see the total picture of what is happening within each region, or between regions, over time.  That kind of information is only provided in the “Inter-Area Movements” table. 

The most important and unique feature about the “Inter-Area Movements” table is that it excludes intra-regional movements of petroleum between countries within each of the regions.  This is vital if you want to truly track the inter-regional movements. 

The “Imports and Exports” table is also useful in that it provides a break down of crude oil and products imports and exports, from which I think I can make some reasonable estimates of the inter-regional movements of these quantities, as I explain below.

BP does not provide a public archive of its past BP Reviews, but fortunately, ASPO international has provided a freely available archive of the microsoft excel files corresponding to the 2002 to 2010 reports, and a pdf file format of the 2001 report.

2)  EIA International Energy Statistics
The USA’s Energy Information Agency’s International Energy Statistics Division was another useful source of statistics on petroleum and crude imports and exports. 

Unfortunately, as I reported several months ago in There will be no Epilogue, the USA can no longer afford to gather and report international energy statistics, and so the international export and import data stops around 2008/2009. 

As an aside, if these data are of interest to you, then you might want to download them now before they disappear altogether or the EIA becomes a pay-per-view site.  Similarly, my feeling is that the BP review’s days are numbered.  While BP’s risk of bankruptcy seems to have faded for now, it is likely only one more Deepwater type of oil spill disaster away from vanishing.


Methods of Analysis
Although this might seem like a simple exercise in data collation, some detailed analysis is needed to put these data condition in order to consistently and properly compare one year to the next and to understand what is being compared.

The BP review’s “Inter-Area Movements” and “Imports and Exports” tables are most decidedly not set up for a year by year comparison. 

The problem is that the regions being reported from year to year are almost continuously changing.  In the 2001 BP review, the year 2000 data separately reports petroleum movements for Western Europe and Central Europe, but in subsequent years just reports Europe.  The 2001 review takes the Asia-Pacific and divides and separately reports “Australasia” (Australia and New Zealand) Japan and China and reports the reminder as “other Asia Pacific.”  But by the 2007 review, the 2006 data further divides out Singapore, and, in the 2009 review, the 2008 data further divides out India.  For all of the available BP reviews, Africa is divided into North, West and East Africa and the trade movements into and out of Africa as a whole, with the exclusion of intra-regional movements between the North West and East, is not reported.  Similarly, the USA, Canada and Mexico are separately reported, and trade movements into and out of North America with exclusion of intra-regional movements between these three countries, is not reported. 

Based on these considerations, I decided to define nine regions and to correct the “Inter-Area Movements” and “Imports and Exports” tables for intra-area movements, as needed: 

1) North America (NA)
2) South America (SA)
3) Europe (EU)
4) former Soviet Union (FS)
5) Middle East (ME)
6) Africa (AF)
7) China (CH)
8) Japan (JP)
9) Asia-Pacific remainder (APr)

As a separate region, Japan is somewhat smaller in area and population compared to the other regions, but, its data is consistently divided out (unlike India and Singapore), and its trade movements of oil are comparable in size to the other regions.  Plus comparing the trade movements of Japan and China over the past decade makes for an interesting story as, you will see.

Corrected Inter-Area Movements
Some of these regions (SA, FS, ME, CH, JP) are as-reported in the BP review, and therefore are assumed to properly exclude intra-region trade movements between the countries with each region.

For the “Inter-Area Movements” table it is straight forward to correct the combining of sub-regions or countries to form my larger regions, by simply adding the sum of the Exports or Imports of the countries now being combined into one value.  For instance, using the data from the 2011 review, NA exports to SA will equal the sum of US, Canada and Mexico exports to SA (from the green box in the “Inter-Area Movements” table shown above) to each of these countries:

NA exports to SA = (Ex US to SA) + (Ex Canada to SA) + (Ex Mexico to SA)
NA exports to SA = (769 kbd) + (2 kbd) + (30 kbd)
NA exports to SA = 800 kbd

and, the sum of imports from SA (from the blue box in the “Inter-Area Movements” table)

NA imports from SA = (Ex SA to US) + (Ex SA to Canada) + (Ex SA to Mexico)
NA imports from SA = (2211 kbd) + (91 kbd) + (26 kbd)
NA imports from SA = 2327 kbd

where kbd is defined as thousands of barrels of petroleum per day

Analogous corrections can be made for APr, AF and for EU region, as needed for each of the “Inter-Area Movements” tables supplied in the BP reviews from 2000 to 2010.

Corrected Crude and Product Imports and Exports
For my newly defined regions (NA, EU, AF, APr) the total Crude and Product imports reported in the “Imports and Exports” table cannot be the simple sum of the formerly separate countries now being combined, because this would not exclude intraregional trades.  Further corrections are need.

The data in the “Inter-Area Movements” can provide such a correction, of sorts.  Specifically, the “Inter-Area Movements” table provided the total intra-regional flows for each of my newly defined regions.  For NA, for example (from the red box in the “Inter-Area Movements” table):

Intra-regional Exports = (Ex US to Cd  & Mx) + (Ex Cd to US & Mx) +  (Ex Mx to US & Cd)
Intra-regional Exports = (124 kbd+477 kbd) + (2532 kbd+7kbd) + (1280 kbd + 32 kbd)  
Intra-regional Exports = 4452 kbd

Intra-regional Imports = (Im US from Cd  & Mx) + (Im Cd from US & Mx) +  (Im Mx from US & Cd)
Intra-regional Imports = (2532 kbd+1280 kbd) + (124 kbd+32kbd) + (477 kbd + 7 kbd) 
Intra-regional Imports = 4452 kbd

Of course, the Intra-regional exports and imports are equal to each other, and these represent the correction factor that needs to be applied to the “Imports and Exports” table, as further explained below.

First, we need to recognize that for the “Imports and Exports” table, the sum of Crude and Product exports (or imports), is equal to the total petroleum exports (or imports) as reported in the “Inter-Area Movements.”  For example, for 2010, the USA's total exports in the “Inter-Area Movements” table (orange box in the “Inter-Area Movements” shown above) equals 2154 kbd.  For the USA, Crude Exports and Product Exports reported in the “Imports and Exports” table equal 28 kbd and 2126 kbd, respectively (orange box in the “Crude and Products Imports and Exports table shown above), which sums up to 2154 kbd.

Similarly, for NA the gross sum of Crude and Product exports equals the sums these respective amounts for USA, Canada and Mexico, which for 2010, equals 6292 kbd (violet box in the “Crude and Products Imports and Exports table shown above).  But, we know from the above calculations, that 4452 kbd of this gross sum is due to intra-regional exports.  Therefore, the sum of inter-regional total petroleum exports equals gross exports – intra-regional exports, which in this example is equal to 6292 kbd minus 4452 kbd, or 1840 kbd. 

In other words, of the gross total exports derived from the crude and product export data in the “Imports and Exports” table, only 29.2 percent (i.e., 100x1840/6293) of that corresponds to inter-regional export from NA to elsewhere in the world.  In my analysis to follow in the subsequent posts in the series, I have assumed that this correction factor can be equally applied to estimate the true inter-regional exports of crude and products.  For instance, for NA in 2010, the sum of gross crude and product exports equal 3381 kbd and 2912 kbd, respectively.  I have corrected these values by multiplying each of them by 0.2925 to derived estimate inter-regional crude and product exports of 988 kbd and 852, respectively. 

Analogous correction factors have been estimated and applied to NA as well as the APr, AF and EU regions, as needed for each of the “Imports and Exports” tables in each year of the BP reviews from 2000 to 2010, as needed.

Gross Export and Import Trends from the EIA
The Import and Export data provided by the EIA are more comprehensive than the BP data in that they go back to 1986 and report data for nearly every country in the world.  Therefore it is possible to take the sum of the imports and exports for all of the countries that comprise the corresponding to NA, SA, EU, FS, AF, ME and APr regions identified above.  In fact, except for APr, the EIA provides the sums for these regions already, so it is a relatively simply matter to calculate APr by subtracting the values from the EIA’s “Asia & Oceania” region which corresponds to BP’s “Asia Pacific” region.

But, the EIA import and export data for these regions, is that as the simple sum of the individual countries in each region this is this is the same as what I called the “gross” imports or exports for NA, that is, imports and exports without exclusion of intra-regional movements between the countries within the region. 

As such the EIA data which includes intra-regional and inter-regional movements, presents an interesting counterpoise to the BP data, which specifically excludes intra-region movements.  Indeed for several regions like SA ME and FS the EIA data provides the only estimate of gross exports or imports, since the BP data totally excluded this.  Even the other regions, like APr or AF, don’t give estimates of the total gross imports or exports because the trade within e.g., West, East and Southern Africa or with other Asia Pacifica have presumably excluded any intra-regional trade occurring within these regions. 

I say interesting counterpoise, because the EIA and BP data set overlap in time from 2000 to 2008. The differences between the EIA’s gross exports and imports, and BP’s inter-regional exports and imports should provide insight into the intra-regional imports and exports for each of these regions. 

------------------
Alright, my house keeping chores are out of the way, and I am ready to present my results.  I’ve decided to start with a broad overview of global trends for the global flows of petroleum imports and exports.  As you will see these results show that both gross and inter-regional global imports and exports of petroleum and crude oil actually peaked in about 2007.   See you next time.