Estimating the End of Global Petroleum Exports Part 2 Regional production, consumption and export trends for ME, AF and AP

Here, in part two, I summarize the production, consumption and export trends for three of the seven regions I studied: Middle East (ME), Africa (AF) and Asia-Pacifica (AP). 

Middle East

Figure 1 presents the reported production, consumption and net exports rates (blue, red and green open circles respectively) and the corresponding NLLS best curves (solid lines with the same respective colors ).  According to the BP data set, the largest 2009 producers in the ME include, in order, Saudi Arabia (40% of all production in the ME), Iran (17%) and UAE (11%), with Iraq and Kuwait close behind UAE. The largest two consumers are Saudi Arbia (36%) and Iran (24%). 

Production rates in the ME peaked first in 1977 (8.2 bbs/yr), and then again in 2008 (9.5 bbs/ys).  To model this, I separately fit two Hubbert equations to the 1965-84 time span and to the 1985 time span.  In contrast, consumption rates in the ME are well modeled using a single Hubbert equation fit to the entire 1965-2009 time span. 

The best fit parameters of Qo, Q∞ and the rate constant, "a," are summarized in Table 1 below:

Table 1 summary of best fit parameter for production and consumption for ME
	Qo (bbs)	Q∞ (bbs)	a (yr-1)
Production 1965-84	11	150	0.22
Production 1985-2009	66	455	0.080
Consumption 1965-2009	6	221	0.056

For production, the best fit parameters are consistent with a very rapid, 22 percent per year increase in the production rate from the mid 60s to the mid 80s, followed by a slower yearly increase (8 percent per year) from the mid 80s to present.  The Hubbert equation best fit suggests that the production rate peaked in about 2007-2008.

For consumption, the best fit parameters are consistent with a steady 5.6 percent per year increase in consumption rate.  The Hubbert equation best fit suggests that the consumption rate will not peak until about 2027-2028.

The reported export rates have two maxima in 1977 (7.6 bbs/yr) and 2009 (7.1 bbs/yr), reflecting the two maxima in production.  Due to increasing regional consumption, despite the higher rate of production in 2009 compared to 1977, the second peak in reported exports in 2009 was lower than the reported peak export rate in 1977 (7.6 bbs/yr).  The best fit export trend predicts peak net exports occurring in 1976 (7.5 bbs/yr) and 2004 (6.9 bbs/yr). 

Based on the best-fit production and consumption trends, I estimate that net exports from the ME will reach zero by 2036. 

Africa

Figure 2 presents the reported production, consumption and net exports rates and the corresponding NLLS best curves for AF. 

The bulk of regional production 2009 was from Nigeria (21%), Algeria (19%), Angola (18%) and Libya (17%).  Current political events in Libya will undoubtedly decrease the Libyan production dramatically for at least the next few years.  The three largest consumers were Egypt (23%), South Africa (17%) and Algeria (11%).

Production rates in AF follow three distinct trends which I modeled using three separate Hubbert equation fits to 1965-82, 1983-2002 and 2003-2009.   Similar to ME, consumption rates in the AF are well modeled using a single Hubbert equation fit to the entire 1965-2009 time span

The best fit parameters of Qo, Q∞ and the rate constant "a" are summarized in Table 2 below:

Table 2 summary of best fit parameter for production and consumption for AF
	Qo (bbs)	Q∞ (bbs)	a (yr-1)
Production 1965-82	6.5	54	0.17
Production 1983-2002	28	155	0.074
Production 2003-2009	18	66	0.23
Consumption 1965-2009	4	74	0.059

For production, the best fit parameters are consistent with a rapid stage of annual production increasing at 17 percent per year from 1965-1982 and then a slower stage of production increasing at 7.4 percent per year for 1983-2002, and, the most recent stage of extremely rapid production increasing at 23 percent per year since 2003.  The Hubbert equation best fit suggests that production rates peaked in about 2007 and then will be in rapid decline if the 2003-2009 production rate trend  continues.

Similar to the ME, for consumption in AF, the best fit parameters are consistent with a steady 5.9 percent per year increase in consumption rates.  The Hubbert equation best fit suggests that consumption rates will not peak until about 2013.

The reported export rates follows a complex time course, reflecting the three different stages of production rate increase against that backdrop of steadily increasing consumption.  The peak in reported net exports from AF was reached in 2007 at 2.7 bbs/yr, which is well modeled (same year and amount) by the best fit export curve.

As noted above, regional net exports will rapidly decline if the recent production trend continues.  Based on the projected production and consumption trends, I estimate that net exports from the AF will reach zero by 2018. 

Asia-Pacific

Figure 3 presents the reported production, consumption and net exports rates and the corresponding NLLS best curves for AP. 

In 2009, China was the largest regional producer (47%) followed by Indonesia (17%) India (9%) and Malaysia (9%).  Of these four, however, only Malaysia was a slight net country exporter, and, as I mentioned in part 1, Malaysia is now interested in importing oil from Iran.  The top regional consumer, of course, is China (33%), followed by Japan (17%), India (12%) and South Korea (9%). 

Both of the reported production rates and consumption rates for the AP region have two distinct stages corresponding to 1965-82 and 1983-2009.  Accordingly, I modeled production and consumption using two separate Hubbert equations. 

The best fit parameters of Qo, Q∞ and the rate constant "a" are summarized in Table 3 below:

Table 3 summary of best fit parameter for production and consumption for AP
	Qo (bbs)	Q∞ (bbs)	a (yr-1)
Production 1965-82	1.1	34	0.22
Production 1983-2009	46	213	0.075
Consumption 1965-82	7.3	91	0.17
Consumption 1983-2009	51	531	0.073

For production, the best fit parameters are consistent with a rapid stage of the production rate increasing at 22 percent per year from 1965-1982 and then a slower stage with the production rate increasing at 7.5 percent per year from 1983-2009.  The Hubbert equation best fit suggests that the production rate peaked in about 2006.

For consumption, the best fit parameters are also consistent with a rapid stage of the consumption rate increasing at 17 percent per year from 1965-1982 and slower stage of consumption rate increasing at 7.3 percent per year from 1983-2009.  Still, the annual percent increase in consumption rate is in the more recent stage much faster than the ME and AF (30% and 25% greater rates of increase, respectively). 

In fact, for the seven regions that I studied in this series, AP has the fastest present-day increase in consumption rate.  AF is second (5.9%/yr), and ME is third (5.6 %/yr).  (The all-time record annual consumption rate increase, however, is held by EU, at 15 percent per year, back in 1965-82.) If the trend for increasing consumption rates in AP were to continue, the peak in consumption in 2013 (9.7 bbs/yr) would be about 350% greater than the predicted declining regional production rate in 2013 (2.8 bbs/yr).

It is apparent from Figure 3 that the AP region was never a net exporter since 1965, and, this trend is not likely to change.  The predicted peak in the import rate is projected to occur in 2015 at -6.9 bbs/yr (through I represent exports as +ve and imports as –ve).

Next time I will finish my summary of the production, consumption and export trends for the remaining four regions: EU, FS, SA and NA.

Estimating the End of Global Petroleum Exports Part 1 Introduction and Method

I and others have made the fairly safe, and perhaps too obvious, prediction that the effects of an end in global petroleum exports will likely happen much more sooner, and, have a much greater impact than peak oil itself.  

The economies of all developed countries in the world are highly dependent on the readily available “liquid energy” that petroleum provides.  Most developed countries import far more oil than they actually produce domestically.  Consequently, all facets of modern life:  transportation, agriculture, manufactured good (plastics), pharmaceuticals and other medical products, high technology goods, are heavily dependent upon this continued flow of petroleum imports.   I believe that what happens to global petroleum exports will strongly affect the economies of all countries of the world, but especially the economies of the heavy importers, and, your standard of living.  

In a previous series, I performed an export land model analysis for the USA which considered the petroleum export trends of the top ten exporting countries to the USA.   I estimated this to amount to about 57% of the total world supply of exportable oil (as discussed in part 3 of the series). 

In the present series, I have expanded my analysis to consider petroleum production and consumption trends for the entire world to quantify the regional global pool of exports and to predict how this will change over time.

Just to save some people the time, here’s the “tweet: version of my analysis which I will lay out in several parts over the coming days:

My modeling for 7 regions of the world predicts that net global exports

end in about 2030 or 2035, depending on the assumptions I make.

Okay, if there is anyone left whose attention span lasts longer than 140 characters, I’ll explain my methods, reasoning and the implications of my findings over the next few posts.

Data Sources and Data Analysis Methods

As in past articles, all production, consumption and export rates (dQ/dt) are reported in units of billions of barrels per year (bbs/yr).  I used the production and consumption data (production minus domestic consumption) to calculate a “reported” net exports (or imports).

I relied on the petroleum production and consumption data from the BP Statistical Review.   I have ranted previously about the limitations of the publically available data sources.  The BP data set has the advantage of presenting data from 1965-2009, but, it only presents production and consumption data for selected countries, and the production and consumption data often are not for the same countries.  

However, at least the BP data set consistently reports production and consumption data for six global regions: the Middle East (ME), Africa (AF), North America (NA), South America (SA), Asia-Pacific (AP), and a sixth region called, “Europe-Eurasia” (E-E), which BP defines as "all countries listed above under the headings Europe and the Former Soviet Union".  To me the E-E region doesn’t make much sense, as this is a collection of very diverse countries with very different economic and petroleum production and consumption histories.  So, I elected to split E-E into two more historically coherent groups: the Former Soviet Union (FS), which is also presented in the BP data set, and Europe (EU), which I calculated as E-E minus FS (e.g., EU_production = E-E_production - FS_production). 

I fit a logistic model (often referred to as the Hubbert Equation) to the petroleum consumption and production data reported for each of these seven regions (ME, AF, NA, SA, AP, FS, EU).  I used non-linear least squares (NLLS) analysis to obtain the best fit of the logistic equation to these data. I have talked about the details of this type modeling at length in past posts on this blog, and I refer the interested reader to the series: Refining the peak oil rosy scenario for further details.  

As you will see in future posts, for almost all of these regions, to improve the fit, and hopefully the predictability, I often fitted two or more Hubbert Equations to separate time spans of the production or consumption date sets.  Finally, I estimated the future export (or import) trends as the predicted production rate minus the predicted consumption rate for each of these regions.  

My Regionalism and Fungibility Assumptions

There are two assumptions inherent in my model that I want to note up front.  

1) Regionalism assumption: This analysis estimates net exports for several continent-sized regions.  This means that for each of these regions, the internal consumption needs for the region are assumed to be entirely met by internal production within the region before there are any net exports outside of the region.

2) Fungibility Assumption: If there are net positive exports for a region, this analysis assumes that that this portion of exportable petroleum is distributed to the other regions with complete fungibility.   

I realize that this is not exactly how petroleum is circulated in the real world.  For various reasons of refining capabilities, logistical considerations and political reasons, some countries export their oil to other countries that are far outside of their particular geographical region, and countries next door to them, in turn, import oil from other countries far away (think of the consequences of past bans on importing Libyan and Iranian oil).  Other countries might simultaneously export oil to the country next door, and import oil from other countries that are far away (Canada is my favorite example of this). 

Nevertheless, I do think that the regionalism assumption is reasonable, because in general, it should be less expensive to ship oil to the country within a continental region than elsewhere.  Actually, I think that this will be a growing trend, to promote regional political and social stability as exports from outside of a region dry up and become unaffordable.  For example, consider the countries surrounding Saudi Arabia in the ME that are rapidly headed to ex-exporter status: Bahrain, Oman, Syria, and Yemen.  It will be in the best interests of Saudi Arabia and the remaining oil major exporters of the ME to supply these countries with their needed difference in petroleum, to maintain local political stability.  For example, Venezuela already sells its oils at a discount rate to some of its neighbors in Latin America (Latin America's New Petro-Politics).  For example, (not withstanding the NAFTA violation) think of the problems that Canada would face if it choose to export its oil to other countries, and the USA, was unable make up for this in the export global market.  Think of the problems that the USA will face, if Mexico becomes an ex-exporter and it can’t afford to import its oil.  The USA and Canada exporting oil to Mexico? Yes I do think that this is possible.  We shall see soon enough, as Mexico seems to be rapidly headed to net zero exports.  

I also think that the fungibility assumption is reasonable because, despite various bans in the past and present, a valuable commodity like oil will eventually find a buyer.  For example, US and EU sanctions against buying Iranian oil doesn’t stop China and Malaysia from filling in the gap (Is the U.S. Bullying Europe into Cutting Ties with Iran?).  Of course, the oil that China and Malaysia would have bought from non-Iranian export sources get bought by someone else and the ban has little impact on Iran.  That's fungibility.

Okay that’s enough for today, next time I will start to present the results of my analysis of the regional production, consumption and export trends.