Predicting Global and Regional Petroleum Consumption Trends Part 10: remaining Asia Pacific

There is quite a bit of material to cover here, so I will limit my introductory remarks to a reminder that the "remaining Asia Pacific" (rAP) refers the group of countries in Asia, other than the countries of the former Soviet Union (FS), China (CH) and Japan (JP), all of which have been discussed in previous parts of this series.

The remaining Asia Pacific’s production, consumption and net export trends: an ELM analysis

Figure 44 presents the BP review’s reported petroleum production, consumption and calculated net exports rates (dark blue, bright red and dark green open circles respectively) and the corresponding nonlinear least squares analysis (NLLS) logistic equation best-fit curves (solid lines with the same respective colors). 

 

Production rates in rAP were modeled using two logistic equation fits to data in the ranges 1965-1982 and 1983-2011, respectively (solid blue lines).  Similarly, consumption rates were modeled using two logistic equation fits to the data from 1965-82 and 1983-2011, respectively (solid pink and red lines).

The best fit parameters of Q_o, Q_∞ and the rate constant "a" are summarized in Table 10 below:

Table 10: summary of best fit parameter for production and consumption data for rAP
	Qo (bbs)	Q∞ (bbs)	a (yr-1)
Production 1965-1982	0.99	21	0.20
Production 1983-2011	22	95	0.066
Consumption 1965-1982	6.1	81	0.091
Consumption 1983-2011	17	228	0.087

As illustrated in Figure 44, after declining to a local minimum in 1982, rAP’s petroleum production rate slowly increased and then hit a peak of 1.6 bby in about 2000.  This peak reflects a composite of production peaks for several major producers in this region: Indonesia in 1991, Australia in 2001, Vietnam and Malaysia in 2004, and Brunei in 2006.  Since peaking, the production rate has slowly declined, e.g., in 2011 production of 1.43 bby was 88% of the peak rate in 2000.

In contrast to flat to declining production, rAP’s petroleum consumption accelerated after 1982, similar to CH (see Figure 39, Part 9).  Larger countries in this region seeing multifold increases in consumption since 1982 include: India, South Korea, Taiwan Thailand and Vietnam. 

Still, the overall relative increase in production in this region is not quite as dramatic and sustained as CH’s relative increase.  Part 9, showed how CH’s consumption rate from 1982 doubled two times by 2004, and might double again by 2016, if it can find the oil to import or produce domestically (which I double).  The consumption rate in rAP doubled from 1982 to 1994, but, since 1994, has risen by “only” another 70 percent.  The deceleration is also indicated by the best fit of the logistic equation to the consumption data, suggesting that consumption might be at or near peak.

Predicting Petroleum Export Rates from rAP to other Regions

Figure 45 shows the relationship between petroleum production rates and export rates for rAP, as already worked out in my previous study from a few months ago.   This figure is the same as Figure 9 in Part 2 of “Relationship between Petroleum Exports and Production.”

Despite overall production rates peaking in 2000, the trend is for increasing exports as a percentage of production (solid regression line, r²=0.52).  However, this overall positive trend is combination of declining exports to JP and NA, and, increasing exports to CH, EU, SA and AF.

Figure 46 shows the predicted absolute regional exports from rAP to the other regions.  These plots are based upon the combination of the production rate trends shown in Figure 44 and the export trend lines shown in Figure 45.

According to this prediction scenario, rAP’s absolute total exports (black line) will peak at about 0.7 bby in 2012-2013.  About 43% and 22% of rAP’s total exports in 2011 went to CH and JP, respectively (compared to 35% and 48% respectively in 2000).  Exports to JP are on trend to end in about 2026.  Exports to CH are predicted to peak in about 2015 and decline thereafter.

Predicting Petroleum Import Rates to rAP from other Regions

Figure 47 shows the sum (black line), and individual import contributions, predicted for each of the other eight regions, to rAP.

This figure reminds me of the analogous figure for JP (Figure 37 Part 8) in that import are highly reliant upon ME.  Of the 3 to 4 bby in petroleum imports to rAP from 2000 to 2011, 80 to 72 percent came from the ME.  That’s about the same as the proportion of JP’s imports coming from ME.  What is different is that imports from the ME to rAP have been going up over the last decade while imports from ME to JP have been going down.

It should be no surprise that I am predicting rAP’s total future imports to be substantially defined by what is expected for ME’s future export trends.  Because I am expecting ME’s exports to at peak now and to decline in the near future, I expect rAP total imports to peak in 2012-2013 and decline thereafter. 

Several other regions export relatively smaller amounts (<0.4 bby) of petroleum to rAP (AF, CH, SA, EA, JP, NA) and these add up to about 0.8 to 1.1 bby, with various predictions of ebbs and flows from these regions over the next 50 years, as illustrated in the lower portion of Figure 47.

Predicting Consumption Rates for rAP based on the PIE analysis

I applied my normalization to rAP in the same manner as done for NA, SA, AF, EU, JP and CH.  For rAP, the average calculated consumption rate, based on the summation of production plus imports minus exports for the 2001-2011 time range, was 0.151 ± 0.111 bby lower than the reported consumption rate for rAP as reported in the BP review.  Therefore my normalization for CH consisted of adding 0.151 bby to the predicted future consumption rate and adjusting total net exports downwards by this same amount.  And, like the other regions, I did not attempt to distribute this correction proportionally among the individual absolute exports and absolute import to and from each of the other regions.

Figure 48 shows the production, consumption and net export data, and corresponding best fit curves, the later two now shown as dashed lines.  Added is the predicted net export (light green solid line representing total absolute exports minus total absolute imports with the -.151 bby correction) and consumption (blood red solid line) rate prediction curves, based on my PIE analysis (exports minus imports with the 0.151 bby correction). 

The results presented in Figure 48 suggest that, if rAP’s production rate follows the decline trend predicted by the logistic equation best fit (solid blue line), and rAP’s export and import rates continue along the trend lines shown in Figures 46 and 47, respectively, then the predicted total net export rate curve (solid light green line) reaches a peak negative value of about -3.4 in 2013, corresponding to peak net imports, after which, net imports start to decline. 

My PIE and the ELM analysis are in good agreement, giving very similar prediction curves for rAP’s consumption.  Apparently for the ELM analysis there is enough deceleration in the reported consumption data for the NLLS best fit of the logistic equation to predict a present peak and decline, in agreement with the PIE analysis, which only looks at production, import and export trends (see Part 1 for a reminder).

The rAP’s combination of trends peaking net imports and decreasing domestic production, leads to my prediction that rAP consumption rate is peaking now at about 4.8 bby in 2010-11 and consumption will decline thereafter.  Essentially, declining domestic consumption and the prediction of declining imports from its primary source, the ME, set rAP on a downward trend in the future. 

At least the downwards decline is fairly mild, at first.  From 2011 to 2021, consumption is predicted to decline -13%.  But the decline in consumption picks up after that (as imports from ME and domestic production both go into faster decline) with a 33% decline from 2021 to 2031, and then 43 % decline from 2021 to 2031.  Overall from 2011 to 2041, consumption is predicted to have declined by 67%. 

Summary & Conclusion

The remaining Asia-Pacific regions has seen substantial increases in petroleum consumption for the past three decades, but the trends in this analysis suggest that further increases are unlikely.  After domestic production peaked in 2000, the last decade’s increase in consumption was fully supported by increased imports, and mainly imports from the Middle East.  But, as I pointed out in Part 2 the Middle East’s production has likely peaked and it own domestic consumption is increasing.  This means that the remaining Asia-Pacific regions can no longer expect continued increasing imports from Middle East. 

At least the remaining Asia-Pacific region’s predicted consumption decline rate for the next decade will be a mild -1.3 %/y.  This is much milder than expected consumption decline rates in Japan (-5.2 %/yr), Europe (-3.2 %/yr) or North America (-2.2 %/y) for the next decade and beyond.  But, by about 2021, as the decline rate in imports from the Middle East, and, domestic production rate both steepen, the remaining Asia-Pacific’s consumption rate decline is predicted to be in the range of -3.3 to -4.3 %/y.

A look at the net importing regions

Now that I have finished my survey of the five net importing regions (NA, EU, JP, CH, rAP), and in fact, completed my nine-region analysis, I want to step back and take a broader look at these net importing regions divided into major groups:  the developed world (NA, EU and JP), and, developing Asia (CH and rAP).  I am painting with a very broad brush stroke here, so apologies to Australia and New Zealand for being lumped into developing Asia. 

Looking at these two net importing groups gives us an opportunity to explore and test the concept of what Jeff Rubin refers to a “zero-sum world.”  The concept as I am applying it here is fairly simple: in a world of a static pool of oil available for exports, increasing petroleum consumption in one net importing region due to increase oil imports means decreasing imports in another net importing region. 

Does the present study provide any support for this? 

Yes, it does somewhat, although some other trends are more complicated than that. 

Figure 49 shows the sum of imports from each the net exporter regions (ME, FS, AF and SA) to my two net-importer groups: (NA EU JP; solid lines) and (CH rAP; dashed lines)

Figure 49 shows that from 2000 to 2011, there was a substantial decline in imports from ME to the NA EU JP groups (turquoise blue solid line) and a concurrent similar magnitude increase in imports from ME to the CH rAP group (turquoise blue dashed line).

That is, NA EU JP’s loss of oil from the ME appears to be CH rAP’s gain.  A zero-sum world.

Sort of the same “zero-sum” scenario is repeated for SA’s exports although the trends are less convincing.  From 2000 to 2011 there have been flat-to-slightly declining imports from SA to the NA EU JP group (lime green solid line), but, a concurrent larger increase in imports from SA to the CH rAP group (lime green dashed line).   So “zero-sum” is not as convincing because the CH rAP group gained substantially more imports from SA since 2000, than the decline in imports to the group NA EU JP.  In other words, SA was able to support oil imports to NA EU JP at a flat-to-slightly-declining rate while at the same time, greatly increasing imports to CH and rAP.  The predicted future import trends suggests that perhaps the “zero-sum” trend will become clearer for SA.

Something other than a “zero-sum” scenario has been happing with respect to FS’s and AF’s imports to these two groups.  I will call it “win-win,” because imports to both net importer groups has been going up over the last decade. 

Imports from FS and AF to the NA EU JP group (solid purple and brown lines, respectively) both went up substantially since 2000—in fact, slightly more than enough to offset the declining imports from ME and SA.  However, imports from FS and AF to the CH rAP group (solid purple and brown lines, respectively) also went up substantially since 2000, thereby supporting even greater consumption rate increases in this group. 

Figure 50 shows the consumption data an predicted future consumption (from the PIE analysis) added all together for the two net importer groups.

As you can see despite the declining imports from ME and SA, the NA EU JP group still managed a slight increase in consumption rate in the last decade, since 2000.  This is due to the increased imports from FS and AF that offset the declines from ME and SA and then some.  However, imports appear to have peaked in 2007, at about 8.9 bby, and have gone down since then.  Imports are predicted to steeply decline for the next two decades.

In comparison, imports to the CH rAP group dramatically increased since 2000—this is due to increased imports from all four of the net importer ME, FS, AF and SA.  Total import increases are predicted until 2015, after which a mild decline in imports is predicted.

The predicted decline rate in imports to the NA EU JP group is so steep, that in about five years, 2018, it crosses below the milder predicted import rate to the CH rAP group. 

Of course, a major goal of this study was to get some insight into what consumption rates might look like in view of the predicted domestic production, import and export trends—my so-called PIE analysis—for each of these nine regions.  We know from Part 6, Part 7, Part 8 that domestic production in NA is flat, is declining in EU, and is substantially non-existent in JP.  That steep decline in imports predicted for the NA EU JP group gives us a big hint about what’s in store for consumption rates for these regions going forwards.    

Here it is in Figure 51, the predicted total consumption rates of the NA EU JP groups versus to CH rAP group. 

In view of the above considerations, it is no surprise to see that, over the next 20 years, the predicted petroleum consumption rate decline for the NA EU JP group is much steeper than for the CH rAP group.  For instance from 2011 to 2031 petroleum consumption in the NA EU JP group is predicted to decline from 14.9 bby to 6.2 bby, a 58%  (-8.7 bby) decline.  In comparison over the same period, the predicted decline in consumption for the CH rAP group is from 8.2 bby to 6.1 bby, a 26% (-2.1 bby) decline.

Interestingly, by 2029, the NA EU JP group and CH rAP group are predicted to have nearly the same consumption rate of about 7.0 bby and 6.8 bby, respectively. 

What kind of world will that be in 2029?

------------------------------------------

Well, at long last, I have completed my nine region analysis.  I will linger with the results for the next post or so, to explore the economic and population trends that these predicted consumption trends might imply (hint: I think that looking at per capita consumption could provide some interesting insights).

Predicting Global and Regional Petroleum Consumption Trends Part 9: China

Some analysts, looking at China’s (CH) stunning economic growth over the past 20 years, expect this trend to simply continue, thereby resulting in China overtaking the USA, to become the country with largest economy in the world.  For instance, the OECD estimates that China's economy will over take the combined economies of the Euro-Zone by 2013, and, then overtake the USA by 2017.  China's economy to overtake US in next four years, says OECD.  Similarly, based upon recent past growth rate, inflation rate and Yuan appreciation rate trends, the Economist predicted that China’s GDP will overtake the USA’s by 2018.  Indeed, the Economist points out that China has already over taken the USA in manufacturing output, patents granted to residents, car sales and total energy consumption, and, that China’s oil consumption would overtake the USA’s in 2021 (The Dating Game).  The EIA notes that China was the world's second-largest consumer of oil and liquids in 2011, and, second-largest oil importer, trailing only the USA in both categories (Economic growth continues to drive China's growing need for energy)

Some economic analysts, however, have a more pessimistic view, seeing signs that China’s export market is weakening, leading to the suggestion that at least the growth rate of China’s economy will be slower in the future if not declining (China Economy Heading for ‘Hard Landing’ as Exports Falter, Shilling Says).   

Still fewer economists, like Jeff Rubin, recognize that China’s continued economic growth will depend upon oil, in particular, oil imports (BP Report Shows Economic Growth Still Depends on Oil), and, that those oil imports will mostly come from sources that used export to the USA (Where Will China Find the Oil to Power its Economy?). 

Rubin’s view is that we are in a “zero-sum world (12 min).”   In the present context, zero-sum means that if exports increase to one region (e.g., China and the remaining Asia Pacific Regions) then exports to another region (e.g., Japan, North America, Europe) must go down.  This is a view that this series supports to some extent, and I will try to point where it does, in this, and the next post.

As you will see, China’s imports have indeed been increasing from several regions, some of which have traditionally been strong exporters to North America.  These increasing import trends should support China’s economic growth for the next five to ten years, albeit at a slower rate than seen in the last decade.  However, if oil consumption and the economy are tightly linked, then, as petroleum exports to China’s start to decline, the prospects for further economic growth also will decline.

China’s production, consumption and net export trends: an ELM analysis

Figure 39 presents the BP review’s reported petroleum production, consumption and calculated net exports rates (dark blue, bright red and dark green open circles respectively) and the corresponding nonlinear least squares analysis (NLLS) logistic equation best-fit curves (solid lines with the same respective colors). 

Production rates in CH were modeled using two logistic equation fits to data in the ranges 1965-1982 and 1983-2011, respectively (solid blue lines).  Similarly, consumption rates were modeled using two logistic equation fits to the data from 1965-82 and 1983-2011, respectively (solid pink and red lines).

The best fit parameters of Q_o, Q_∞ and the rate constant "a" are summarized in Table 9 below:

Table 9: summary of best fit parameter for production and consumption data for CH
	Qo (bbs)	Q∞ (bbs)	a (yr-1)
Production 1965-1982	0.24	12.3	0.24
Production 1983-2011	36.2	371	0.026
Consumption 1965-1982	0.22	10.1	0.26
Consumption 1983-2011	7.1	485	0.074

As illustrated in Figure 39, after declining to a local minimum in 1982, CH’s production rate has been slowly increasing over the last 3 decades.  In 1992, the production rate of 1.04 bby was 33% greater than the production rate of 0.78 bby in 1983.  By 2002 the production rate of 1.22 bby was up 18% compared to 1992, and, in 2011, the production rate of 1.5 bby was up 22% compared to 2002. 

However, this growth rate in production of 2-3 %/yr was not enough to keep up with the growth rate in consumption, and 1993, CH’s consumption crossed-over its domestic production to make CH a net importer of petroleum.

CH’s growth rate in petroleum consumption has been truly breath-taking.  Consumption was 0.58 bby in 1982.  It took 13 years, 1995, to double.  Then, in less then 9 years, 2004, it doubled again.  If my logistic equation best fit (red line) were to accurately model future consumption, then consumption would double again in 12 years, in 2016, to 4.6 bby and then double again to 9.4 bby in 2030. 

For reference, according to the BP review, the US’s consumption rate was 6.9 bby in 2011.  But, as I pointed out in Part 6 of this series, North America’s (NA) consumption rate is in decline, and that includes declines in US consumption.  So, if the trends in the US CH and were to continue, and the oil export were available, it is not to hard to see how CH could overtake the USA in consumption be the early 2020s.  Indeed CH could overtake all of NA by the mid-2020s, under this scenario.  I think this this type of extrapolation are what causes the QECD and The Economist to make the predictions that I outlined above, that CH would soon overcome the US in oil consumption and economy (e.g., GDP).

I think such scenarios are a fantasy however. 

The problem with this type of scenario is revealed by the size of the solid green line in FIG. 39 going forward—this is the predicted net imports of petroleum that CH would have to have to provide the projected consumption rate, as represented by the solid red line.  In 2006, CH’s imports of about 1.3 bby were about equal to CH’s domestic production rate in that year.  By 2011, imports exceeded 2 bby, while domestic production was 1.5 bby.  By 2020, CH’s imports would have to double to 4 bby and domestic production increase to 1.7 bby, to support the projected consumption rate.  Actually, the imports would have to increase by more than 4 bby, since CH still exports petroleum to other regions, and, the trend is for these exports to increase going forwards. 

Unfortunately, the scale CH’s projected imports and domestic production are not on trend to provide this level of petroleum for domestic consumption.   Production, the blue line is only gradually increasing, and so the bulk of petroleum to support the projected increase in consumption would have to come from increased net imports. 

Let’s look at CH’s predicted exports and imports trends in detail. 

Predicting Petroleum Export Rates from CH to other Regions

Figure 40 shows the relationship between petroleum production rates and export rates for CH, as already worked out in my previous study from a few months ago.   This figure is the same as Figure 8 in Part 2 of “Relationship between Petroleum Exports and Production.”

As illustrated, CH’s total exports (black Xs) are only about 10 percent of CH's total production, but, there is a positive trend (solid regression line, r²=0.44) for an increasing proportion of domestic production to be exported.  Within that overall increasing trend, CH’s exports to NA and JP are going down, and, exports to the remaining Asia-Pacific region (rAP) and South America (SA) are going up.  By far, most of CH’s exports goes to it near neighbors in the rAP, and as discussed in my previous series in Part 2, most of these exports are petroleum products exports.  

Figure 41 shows the predicted absolute regional exports from CH to the other regions, based upon the combination of the production rate trends shown in Figure 39 and the export trend lines shown in Figure 40.

According to this prediction scenario, CH’s exports will steadily increase, going mostly to rAP and SA, and to much lesser extents, EU and AF, while exports to NA and JP are already over.

Predicting Petroleum Import Rates to CH from other Regions

Figure 42 shows the sum (black line), and individual import contributions, predicted for each of the other eight regions, to CH.

As already discussed in the earlier parts of this series, exports to CH from ME, FS, AF, SA, and even NA and JP, are all trending upwards.  These trends are reflected by the rapid increase in CH’s total imports increasing from 0.5 bby in 2000 to about 2 bby in 2010.  This is predicted to continue, until production from CH’s major suppliers, AF, FS, rAP, ME, and SA all eventual peak and decline, and along with it, declining exports to CH.  According to this prediction scenario, total imports to CH peak at about 2.5 bby in 2018 and decline thereafter.  There is a pretty broad plateau from about 2013 to 2024, where imports stay above 2.3 bby, but then starts to decline especially as exports from ME to CH go into decline.

Of course, CH’s net imports (imports minus exports) will by less than these amounts, since as discussed in the context of Figure 41, CH domestic exports are predicted to increase over this period.   

Predicting Consumption Rates for CH based on the PIE analysis

I applied my normalization to CH in the same manner as done for NA, SA, AF, EU and JP.  For CH, the average calculated consumption rate, based on the summation of production plus imports minus exports for the 2001-2011 time range, was 0.038 ± 0.076 bby lower than the reported consumption rate for CH as reported in the BP review (interestingly, this is about the same normalization applied to JP’s data).  Therefore my normalization for CH consisted of adding 0.038 bby to the predicted future consumption rate and adjusting total net exports downwards by this same amount.  And, like the other regions, I did not attempt to distribute this correction proportionally among the individual absolute exports and absolute import to and from each of the other regions.

Figure 43 shows the production, consumption and net export data, and corresponding best fit curves, the later two now shown as dashed lines.  Added is the predicted net export (light green solid line representing total absolute exports minus total absolute imports with the -.038 bby correction) and consumption (blood red solid line) rate prediction curves, based on my PIE analysis (exports minus imports with the 0.038 bby correction). 

The results presented in Figure 43 suggest that, if CH’s production rate follows the trend predicted by the logistic equation best fit (solid blue line), and CH’s export and import rates continue along the trend lines shown in Figures 41 and 42, respectively, then the predicted total net export rate curve (solid light green line) reaches a peak negative value of about -2.2 in 2017, corresponding to peak net imports, after which, net imports start to decline. 

CH’s projected slowly increasing domestic production rate causes CH’s consumption rate to peak a few years after the import peak—at about 3.82 bby in 2020. 

That peak consumption rate of 3.82 bby is only about 7% higher than CH's consumption rate of 3.56 bby in 2011.  This would imply a much milder growth rate in petroleum consumption over the remainder of the decade, about 1%/yr, than seen in the past decades.   

The predicted decline in consumption rate in the first decade after 2020 is comparatively mild—about an -11% decrease to 3.4 bby by 2030.  The decline rate picks up steam after that (as imports from ME go into faster decline), declining to 72% of the peak by 2040 and then to 57% of the peak by 2050.  If these predicted production and consumption rate trends continued, then CH’s consumption rate and production rate would cross-over again in about 2049, making it possible for CH to become a net exporter, or alternatively, maintain its consumption rate for a time at about the same level it was at in the mid 2000s.

Summary & Conclusion

My last three posts, discussing petroleum consumption trends in North America (Part 6), the European Region (Part 7) and Japan (Part 8), all illustrate how, over the last decade, exports from the Middle East and other net-exporter regions to these three regions have been going down. 

In the present “zero-sum world,” as Jeff Rubin coins it, it is mainly the decline in exports to these three regions that has supported increased exports to China, and as you will see in the final post of this series, to the remaining Asia-Pacific regions.   

Zero-sum in this context is fairly accurate in the case of the Middle East and South America, because their net exports have been fairly flat over the past decade (see e.g., Part 2 and Part 5 Figures 1 and 17, respectively, green circles).  For other net exporters, like the former Soviet Union region and Africa, however, this is not the case, as net-exports from these regions actually increased.  But I do not expect this increase to continue.  Rather, I expect that exports from the former Soviet Union and Africa will decline as these region's production rates decline. 

The good news for China is that the trends for continued increases in domestic production and imports, even while slightly increasing their petroleum product exports, will make it possible for China to have continued real economic growth, possibly up to the end of the decade--unlike North America, the European Region and Japan.

After 2020, however, the picture looks less rosy for China, as it’s main suppliers of oil start to go into production and export decline.  This probably means that China’s economy will become stagnant in the 2020s and then decline thereafter.  Even then, the predicted declining consumption rates in China of -1.1 %/yr from 2020 to 2030 or -1.4%/yr from 2020 to 2040, and, the possibility of a flattening consumption rate after that (e.g., due to increased domestic production, perhaps from the East China Sea and Western China), suggest a much milder decline scenario than what the decline trends look like for Japan (-5.2 %/yr), Europe (-3.2 %/yr) or North America (-2.2 %/y) all of which are on-going presently, and, will likely continue for the rest of the decade and beyond.   

------------------------------------------

Next time, I will cover the final region in my 9-region analysis: the remaining Asia-Pacific region.