On 2012 Mar 07, The U.S. Energy Information Administration released a report on the history of U.S. liquid petroleum products (crude oil, gasoline, kerosene, …) highlighting the positive fact that exports are leading imports for the first time in 6 decades. However, our analysis of EIA data uncovers a disconcerting trend. Exports show a huge jump about 1980 (during the Oil Embargo), ratcheted another jump in the late 1980’s ( lead-up to Desert Storm), and have blasted upwards during the past 7 years to reach its current unprecedented 5O+% of our crude oil production. This is about 27% of our total liquid petroleum production.
Added after initial publish date : The point of this post is not that we are exporting our petroleum, but that exports are growing way out of line compared to our usage, or even to our imports! The data (below) show that similar export jumps happened 3 times in our history, when our politicians seemed to promote resource exploitation for someone’s personal gain.
This post is an update to US Petroleum Exports are Huge. LastTechAge is presenting it because this startling chart has been taken off-line at least once in the past week, and because internet data is politically fragile. Here is the original EIA link, also shown in Huge. Here is the PDF of the originalEIA report.
These jumps in exports happened when we as a country were under high stresses and could blame rising prices on other factors. Although there are no names attached to the data, U.S. oil companies control our petroleum inventory. My opinion is that the export jumps represent profiteering actions on the part of those responsible.
Their lead chart is shown as our Fig 1. Click to enlarge. Export data is the boundary line of the top (light brown) data. On a quick scan, exports do not look all that shocking compared to imports. For production, imports and price trends of American crude oil during this time (1949-2011), refer to Fig 1 in Footprints-2 Oil.
There were 3 distinct jump periods when exports exponentiated from a starting value. More details on how the analysis is done are given at the end; in summary, data trends rise exponentially when they follow a straight-line path on a semi-log (log-linear) plot.
Fig 2 is a semi-log presentation of the EIA data. The 3 jump periods are clearly visible as straight lines on this chart.
The doubling time (table below) is the time required for exports to double in size. For example, 2004 exports were 1 M bdp. Exports doubled (exceeded 2 M bpd) about 4½ years later, and should reach 4 M bpd in early 2013 (second doubling). If trends continue, it should reach 8 M bpd (third doubling) in an additional 4.6 years.
While 2 to 4 hour gas lines were the norm, US exports rose like there was no tomorrow. And we thought monster gasoline prices were caused by the embargo.
This first jump period ended abruptly in 1983, when the country started preparing for presidential elections. Hard stops of exponential growth indicate sudden application of some kind of external control.
The second jump period began smoothly in 1986. Although these data do not indicate why petroleum exports jumped, they rose 9%/year right through our Desert Storm war, when we actually needed the oil to support our troops. The rapid rise abruptly stopped a year before presidential elections in 1992.
The third jump period started smoothly in 2004, and is exponentiating with 15% growth in export volumes every year (the jump data form a straight line in Fig 2). This continued smoothly through the 2008 election, rose without pause as gasoline prices have risen, and is still shooting upwards according to the latest data point (2011).
Fig 3 shows the same EIA data but in a normal linear plot, so that we can see with ‘normal eyes’ how well the exponential model works with petroleum export data. The 3 straight lines found in Fig 2 all curve upwards through the data points.
Guesses about why this has been happening will be the topic of future posts. But the fact oil has risen to well over $100/b for WTI quotes and more than $120/b for European Brent Oil quotes probably gives a hint.
Highly robust export growth of American petroleum products.
In 2011, Obama tried to force oil prices down by releasing precious oil from the Strategic Petroleum Reserve but at each release, there was a fairly quick bounce back to higher prices. I was puzzled by this in 2011, had no clue as to why the releases did not work. Current Talking Head broadcasts discuss this problem in light of the Republican push to open the SPR again for additional lucrative sales. Obama just might do so in a vain attempt to help election year prices.
Identification techniques for exponential growth data:
This PDF explains how to identify exponential growth segments in a data set. If you plot the data on a log-linear (semilog) graph, an exponential curve shows up as a straight line, as in Fig 2.
The related doubling time, TDBL, is another way to identify and characterize exponential growth. If export volume grows exponentially, then it will grow by a factor of 2 when the time TDBL passes. TDBL = ln(2)/r. Since r is the rate in decimals, a 15%/yr growth rate means TDBL = ln(2)/0.15 = 4.62 years.
Analyst’s View When I wear my physics hat, I have high confidence that data are exponential when the data points (a) rise along a semilog straight line and (b) growth is > a factor of 10. With my engineer hat (I also worked as technical support in industry), five of the same exceptions in a row trigger the statistical process control (SPC) signal that a significant effect is occurring.
The current (2011) jump has lasted for seven successive years. The 1980 jump was only 4 anomalous data points, but they had the fastest rate, the greatest change, and visually follow a straight path. The 1990 jump lasted for the SPC 5 years. The points scatter about a semilog straight line, and all have been tagged as free exponential expansion, impeded by only their own internal natures. Even if the 1990 jump were curved, it represents a significant change in exports.
This update is meant to complement our previous post by showing the graphical reasons behind the observations, and to generate comment. It is also meant to preserve the public domain information from any possible removal.
Update: 2012 May03 Update to this post: 2012-May03 see comment, 2nd paragraph.
Charles J. Armentrout, Ann Arbor
2012 Mar 20
Have a comment? Click the title of the post, go to the bottom and let us know
Listed under Natural Resources …thread Natural Resources > Oil
Related posts: Click the INDEX tab under the Banner picture, or click [Categories], right