/* Added by TWP, 10/12/2012 */ /* End of addition */

One of the live oaks that bless my home

Saturday, September 29, 2012

The Discrete Charm of Drilling in America

This blog is a companion to "The Discrete Charm of Living at the Peak."

If you follow national media, you are probably convinced by now that in 10-15 years from today, the United States will be producing enough oil to become independent of foreign oil suppliers.  (In most predictions of energy independence, Canada and Mexico are treated as the almost domestic oil suppliers.)

So can the United States of America be dependent only on domestic crude oil production and imports from Canada augmented by Mexico? This scenario is not as nonsensical as it may sound, if (1) the United States continues to destroy demand for petroleum just as it has in the last four years; and (2) crude oil imports from Canada increase dramatically, because Mexico will not be able to export much crude oil in 5-10 years from now.  Since 2008, the U.S. has destroyed demand for 2 million barrels of crude oil per day, which translates into an average annual destruction rate of about 0.5 million barrels of oil per day.

If each year of the next decade the US petroleum demand diminished by 0.5 million barrels of crude oil per day, we would be consuming only 19-5=14 millions barrels of oil per day in 2022.  If we also imported 4 million barrels of oil per day from Canada (1.6 million barrels more than the current 2.4 million barrels of per day), only 10 million barrels per day of demand would be left.  The domestic production of crude oil might increase to 7 - 7.5 million barrels of oil per day from the current rate of 6 - 6.5 million barrels of oil per day, but this increase will be difficult to achieve.  The remaining demand for 2.5 - 3 million barrels of oil per day might be satisfied by switching to domestic natural gas and natural gas liquids.

Sun sets on the Suncor Millennium open pit tar sand mine in Fort McMurray, Alberta, Canada. Source: Robert Kunzig, The Canadian Oil Boom, National Geographic, March 2009.
An increase of 1.6 million barrels of oil  per day in US imports would require Canadians to double current production from the tar sands in Alberta.  Given the dire environmental impacts of such doubling, my proposed Canadian imports solution is iffy at best.
Click on this chart to see it in high resolution.  The US DOE EIA data for crude oil production and active rotary rigs have been put into an Excel spreadsheet by Mr. Dave Room of the US Association for the Study of Peak Oil (ASPO), and plotted in a similar form.
The chart above shows monthly production of crude oil in the United States versus the number of active rigs drilling for oil.  I have assumed that on average a newly drilled well starts producing after a six-month delay. In the chart, the monthly oil production rates are color-coded as deep red for the oldest data, and follow rainbow colors up to deep purple for the newest data.

The monthly oil production and the number of active rigs decreased almost monotonically each year between 1987 and 2000.  The years 2000 and 2001 (hues of blue), showed the largest scatter and a reversal of direction.  Between 2001 and 2012, oil production increased a little with a six fold increase of active oil rigs.

We could assume that going backward in time from the year 2000 to 1987, would result in a substantial increase of oil production with the increasing number of active rigs: the production would increase about 60 percent with the number of rigs quadrupling. Not so between the years 2001 and 2012: the oil production increased less than 20 percent, while the number of rigs grew six-fold.  But reality is even more unsettling.  The newer rigs are 2 - 4 times more efficient than the old ones, so the relative effort of maintaining oil production in the US has increased not just 7 times, but perhaps as much as 14 - 28 times!
Barrels of oil produced in the US per month per active rig vs time.  It is assumed that production from each rig lags by 6 months on the average.  Rig productivity peaked in the year 2000, and has declined with an average exponential rate of 17 percent per year. Data source: Mr. Dave Room of ASPO.
Another way of looking at the same data is shown in the chart above as 25 years of oil rig productivity in the Unites States.  The plot is semi-logarithmic to show exponential decline of rig productivity at the rate of 17 percent per year.  Thus, to maintain or increase our domestic production over the last decade, we have entered the classical Red Queen's race.

The Red Queen's Hypothesis, or "Red Queen Effect"  is taken from the Red Queen's race in Lewis Carroll's Through the Looking-Glass. In Chapter II, the Red Queen said, "It takes all the running you can do, to keep in the same place." The Red Queen Principle can be stated thus: "For an evolutionary system, continuing development is needed just in order to maintain its fitness relative to the systems it is co-evolving with."
'Well, in OUR country,' said Alice, still panting a little, 'you'd generally get to somewhere else—if you ran very fast for a long time, as we've been doing.'
'A slow sort of country!' said the Queen. 'Now, HERE, you see, it takes all the running YOU can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!'
'I'd rather not try, please!' said Alice. 'I'm quite content to stay here—only I AM so hot and thirsty!'

Lewis Carol, Through the Looking Glass, Chapter II. The Garden of Live Flowers.

Well, in OUR country, we run faster and faster to maintain oil production, and we are getting hot and thirsty in North Dakota's Bakken oil play, or in Texas' Eagle Ford Shale and the Permian Basin.  In the mean time, it is also increasingly more difficult to maintain ultra deep offshore production in the Gulf of Mexico, and Shell's foray into the Arctic is proving far more difficult and expensive than expected.

As oil well drilling and completions get faster and more robust, the reservoirs these wells penetrate become less accessible,  less permeable, and less rich in oil at equal speed.  Thus, travel on the road to the future oil independence of these United States feels like Alice trying to outrun the Red Queen.

P.S.  The same trend holds for Texas and natural gas.  The chart below plots the history of average gas well productivity in Texas.  If nothing is done to make future wells more productive (and a lot is being done), an average gas well in Texas will cease production 30 years from now.
Average gas production per well in Texas.  This plot was obtained by dividing the total gas production reported by the Texas Railroad Commission through the number of active gas wells.  Source: Texas Railroad Commission website, accessed on 08/05/2012.
It is obvious that employment in the oil and gas industry must increase dramatically to staff the new difficult and well-intensive projects.  This new increased employment is no longer a boom-or-bust proposition, but is necessary to keep hydrocarbons flowing in the US and elsewhere in the world.

1 comment:

  1. Hi, Mr Patzek
    First time commenting so let me begin by thanking you for this truly inspiring blog that I read regularly.
    Now on this article. There’s something I find somehow confusing in it. It seems to convey the general idea that the more oil we extract from the ground, the more difficult to extract it is the oil left. I agree with that general idea and I think that there are plenty of evidence supporting it. But as I see it the data provided in this article may not be among that body of evidence. I am focusing on the first of the graphs. What it shows, if I am not misinterpreting it, can be understood like that:
    - The oil wells have, as a whole, a natural depletion rate. If you want to keep production at a level you need to drill enough new wells to offset the depletion of existing wells.
    - From 1987 to 2000, the rate of drilling was below that offsetting level and decreasing, so production also decreased, following an almost linear path.
    - After 2001, the rate of drilling increased heavily, but not enough to surpass clearly the offsetting level so production didn’t increased substantially.
    - Going backwards in time through the 1987-2000 path is not possible not necessarily because oil is today harder to find and extract than it used to be, but because it is impossible to reverse the depletion process of the existing wells.
    So a huge amount of drilling may be necessary to go back to 250 MMBO/month as in 1987, but the main reason could be that not enough drilling for oil had been done in the previous years. Oil may be as abundant or scarce as it use to be back in 1987. In other words, we may have to run much faster today not because we have to outrun an ever faster queen but because we have to catch up after several years of slow walk.
    I have tried this very simple, recursive model:
    P(i) = (1-r)·P(i-1) + p·C(i)
    Where P(i) is production in year i, r is the general rate of depletion of the system of wells, C(i) is the number of rigs reaching completion in year i, and p is the average initial production of a new well. With r = 4% annual and p = 0.0056 MMBO/month the model fairly fits the monotonic descent of 1987-2000 (I don’t have the data, my fitting is a general visual one only, but the relationship between p and r is chosen to get that 0.14 MMBO/month-rig displayed on the graph). Then, keep the drilling rate around 150 active rigs for 2 years to allow for the scattered blue points. Now, if you feed the model with an increasing drilling rate up to 1200 in 2012 what you get is still a decreasing production. To get something that fits with the pink points in the graph you can suppose that initial production of new wells in 2003-2012 is something as 0.015 MMBO/month, three times higher than previous p. Or you could also allow for a smaller depletion rate, r (1.5%, three times smaller than the previous one). I don’t know if this crude model can fit with any realistic picture of what’s happening down there but it can explain the data on graph one without any increasing scarcity of oil underground.