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

One of the live oaks that bless my home

Wednesday, November 21, 2012

The Global Oil Peak or a Plateau?

I am about to cover a very serious subject, so please forgive my somewhat formal and unduly precise language.  Since I am talking here about the future of our crude oil-powered civilization, I do not feel too guilty. Besides, you can always stop reading...

The six categories of liquid and solid hydrocarbons in Figure 1 are lumped together into three different combinations in the reports of global liquid fuel production maintained by the Energy Information Administration (EIA).
Figure 1: Click on the image above to see its full screen version. All liquid and solid hydrocarbons and alcohols are grouped into six categories that exhaust the classification of liquids used by the U.S. DOE Energy Information Administration (EIA) in their reports of global fossil fuel production. Note that depending whether heavy oil flows or not at the initial reservoir conditions, it is classified as either a transitional crude or unconventional crude. Solid tar sand bitumen is mined and liquified in surface plants. Kerogen in oil shale is a solid that is either mined and liquified in surface reactors, or liquified in situ using heat.  This figure was drafted by Mr. Erik Zumalt of UT Austin.
These combinations are:
  1. Natural gas plant liquids (NGPLs). NGPLs are those hydrocarbons in natural gas that are separated as liquids at natural gas processing plants, fractionating and cycling plants, and in some instances, field facilities. Lease condensate is excluded. Products obtained include liquefied petroleum gases (ethane, propane, and butanes), pentanes plus, and isopentane.
  2. Lease condensate and crude oil. Lease condensate is a mixture consisting primarily of hydrocarbons heavier than pentanes that is recovered as a liquid from natural gas in lease separation facilities. Lease condensate is lumped together with several types of crude oil that are also classified as Easy Oils, Transitional Oils, and Unconventional Oils. Starting from the second column on the left, these crude oils are the light conventional crude oils, heavy oils, ultra-deep water oils, Arctic oils, tight mudstone ("shale'') oils, ultra-heavy oils, as well as tar sand bitumens and kerogen from oil shales. The ultra-heavy oils are recovered in situ by heat injection, mostly as steam. The bitumens and kerogen must be liquified either at upgraders/refineries or in situ. All of these liquids together make the "Crude oil + lease condensate" curves in Figures 2 and 3.
  3. Other liquids. These liquids lump gas-to-liquids (GTLs), coal-to-liquids (CTLs), ethanol from corn and sugarcane, biodiesel from palm oil and soybeans, and any other liquids that might be used as fuel, i.e., methanol, butanol, etc.
The classification of liquid and solid hydrocarbons is shown in Figure 1, and their production rates are plotted in Figures 2 (by volume) and 3 (by mass translated into energy). Note that if liquid hydrocarbons are reported by mass (the only correct reporting), refinery gains disappear, because they only add volume by decreasing the liquid product density, but they do not change the product mass. 
Figure 2: The rate of global production of liquid fuels by volume in millions of barrels per day. The natural gas plant liquids (NGPLs) plus "other liquids'' (CTL, GTL, biodiesel, methanol, etc., but mostly ethanol from corn and sugarcane) are on top of the crude oil and lease condensate curve. The "Crude oil and lease condensate'' term includes all naturally occurring liquid and solid hydrocarbons shown in Figure 1. Note that the average NGPL density is at most 65% of the average density of 0.84 g/cc of the crudes produced around the world. Thus, volume-based plots are misleading when it comes to the specific energy content of a fuel (energy per unit mass). Simply put, a gallon of one liquid fuel has a different energy content than a gallon of another fuel. All other factors being equal, only the specific energy counts when it comes to driving. Refinery gains of up to two million barrels per day are not shown, because they are a volumetric illusion that does not contribute to fuel energy, see Figure 3. Note that the rate of global production of crude oil and lease condensate has remained unchanged since 2004. Depending on your favorite time scale (years vs. decades) this constant production rate is either a "plateau" of crude oil and lease condensate production or an "oil peak,'' or -- more accurately -- a "crude oil and lease condensate peak.'' Volume data source: EIA's database, accessed on 11/18/2012.
Figure 3: The mass rate (volume rate x density) of global production of liquid fuels, converted into an equivalent energy rate in exajoules (EJ) per year. 1 EJ = 10^18 joules ≈ 10^15 BTUs or quads. The higher heating values (HHVs) of all fuels were used for the conversion from mass rate to energy rate. The natural gas plant liquids (NGPL) and "other liquids'' are now equivalent to the crude oil and lease condensate, because all volume rates have been converted to mass rates and, subsequently, energy rates. The refinery gains disappear altogether, because they do not add energy by mass, but only volume. In fact, refinery gains cost external energy dissipated by thermal cracking and hydrogenation and, thus, diminish the total energy available as fuels. Volume data source: EIA's database, accessed on 11/18/2012. The liquid densities and HHVs are from several sources.
In contrast to the EIA classification described above, both the International Energy Agency (IEA) and BP report different combinations of the liquids and solids classified in Figure 1. In particular, IEA splits all natural hydrocarbon mixtures into conventional and unconventional oils using their own definitions, and BP lumps all natural liquid fuels together, see Figure 4. These different classifications, plus general ignorance of the public and popular media, lead to endless problems and differing interpretations of the same data. Even the names of the respective agencies (EIA vs IEA) are confused and used interchangeably.
Figure 4: The BP oil production data were downloaded from BP's statistics website, as millions of metric tons of oil equivalent (toe, accessed on 11/21/2012). This mass rate of global production was converted into HHV by multiplying it by 41.868 x 10^9 GJ/toe x 1.07 to convert from the standard lower heating value of 1 toe to its HHV, because only HHV can be used to compare fuels with different hydrogen contents. The EIA curve represents lease condensate plus crude oil plus NGPLs, all converted from volumes to HHV using the procedure described in the caption of Figure 3. Notice excellent agreement between these two curves. The IEA data could be purchased for a lot of money, but were not.
If one recognizes that the U.S. EIA's "crude oil and lease condensate'' curve lumps literally every natural hydrocarbon that is not natural gas plant liquids nor synthetic fuels or biofuels, one must conclude that the global production rate of natural hydrocarbons has stalled at the level seen already in 2004, or 8 years ago. Daniel Yergin and IHS CERA call this phenomenon an "undulating plateau of production rate'' others call it an "oil peak.'' The two sides enter into endless debates about whose interpretation is better, but the empirical fact remains: The global rate of liquid and solid hydrocarbon production has stopped growing since 2004. Call this empirical observation by whatever name that better suits your taste, but first please look at Figure 5.
Figure 5: I set up this model of global oil production probably in 1995, or so, and never changed its parameters.  I have only updated the blue data curve, which is a superposition of the old historic data from a variety of sources and the EIA data. By a lucky coincidence, or the Central Limit Theorem, or both, the world production of crude oil and lease condensate has been quite predictable for the last 17 years or so.  I want to point out that there will be future small Hubbert curves for the new Iraqi oil, GOM oil, the Arctic oil, etc., but the fundamentals will not change, just as they are unchanged for the Norwegian sector of the North Sea shown in my earlier post. At the time scale of this chart, the global oil production plateau surely looks like a peak.

Interestingly, to meet global oil demand between 2010 and 2035, IEA reported in 2011 that $10 trillion would be needed, with the upstream sector accounting for 85% of this amount.  IHS CERA reported in 2012, that the upstream costs more than doubled in 2011, relative to the year 2000.  This trend is expected to continue because higher demand for raw materials, equipment, and specialized labor will create shortages.  Most importantly, the global oil and gas industry is moving away from the easy oil to the transitional and unconventional oils, and this means a continuous increase of cost and complexity of the future upstream operations.  In human language, we will have to run ever faster to stand still.  This is yet another definition of peak capacity to maintain oil production at the current level.  In their 2012 Energy Outlook, IEA predicts that in the year 2035, global petroleum production will be 4,000 Mtoe (mega tons of oil equivalent) vs. 3,600 Mtoe produced in 2010.  The global peak oil plateau, anyone?


  1. Hi Tad,

    Very much enjoy your posts which stand as sentinels of reason in the midst of human folly about the subject of oil.

    I have a couple of questions please. Firstly, what is the second, smaller peak, in figure 5? Secondly, according to your best estimate, when will the rollover event occur (permanent decline in C+C) and how will we know it for what it is?

    Best Regards.

  2. Hi Mercury4,

    The small peak had to do with U.S. oil production increasing rapidly until 1971, more oil production from the Gulf of Mexico later, the beginning of the North Sea production in 1969, and the beginning of production from Prudhoe Bay in 1978. Equally importantly, by 1974, the production cuts announced by OPEC in October 1973, started to kick in. These cuts constituted a serious deviation from the assumption of uncorrelated production from many sources that is the basis for emergence of a Hubbert curve.

  3. A better way of looking at the permanent C&C decline, which will happen in several years, is to subtract the actual oil production from the population-based projections and calculate the unrealized oil consumption. By this standard, the peak of global per capita consumption happened in 2005, and we now live in an era of permanent destruction of consumption. The US and EU consume more per capita, have exported their manufacturing elsewhere, and are unable to withstand the high price of oil. Hence the destruction of demand for 1.6 million bbl per day that already happened in the U.S. A similar destruction trend will continue to happen in the future, because the U.S. can no longer compete for expensive oil with China. I do not think that anyone in Washington D.C. understands this simple and unpleasant fact, and you have the Paul Krugman's of this world reporting their dreams from the Keynesian Wonderland.

  4. Pretty! This has been a really wonderful article.
    Thanks for supplying these details. Italy Oil & Gas Report