The Hubbert Peak for World Oil

Theory:

It is widely accepted that oil is a finite resource; there are basic laws which describe the depletion of any finite resource:

These simple rules were first described in the 1950s by Dr. M. King Hubbert, and apply to any relevant system, including the depletion of the world’s petroleum resources.

The rate of production of a natural resource can be plotted on a graph against time. This gives a picture of the lifetime of that resource…

World Oil Production to date
Figure 1
This graph depicts world oil production to date.

The vertical line indicates the probable midpoint of
depletion as identified by Campbell.
(Refer to future scenarios as estimated by Campbell and others.)

It is important to note that the point of maximum production (known as the Hubbert Peak) tends to coincide with the midpoint of depletion of the resource under consideration. In the case of oil, this means that when we reach the Hubbert Peak,

we will have used half of all the recoverable oil that ever existed on our planet.

When viewing a graph of oil production (millions of barrels per day, Mbd) against time (Years), the area under the curve corresponds to cumulative production - the area under a world oil production graph from 1970 to 1980 would give the total amount of oil produced between the years 1970 and 1980. Thus the area under a curve describing the total lifetime of a natural resource corresponds to the total amount of that resource that was ever and will ever be available for production. This quantity is known as the Ultimate for that resource.

Practice:

If we knew the Ultimate for oil production, we would then be able to tell when we had reached the midpoint of oil production - it would simply be the moment when cumulative production (which is known) was equal to Ultimate÷2. As the midpoint coincides with the production peak, we would also know exactly when production was peaking, and hence when it was going to start declining. This would be useful knowledge, because reaching the production peak has serious implications (see later).

Since we won’t know the exact value of the Ultimate for petroleum until we have run out, the next best thing we
can do is estimate it. There are four important concepts which have to be considered when estimating the Ultimate for oil production:

Quite simply, Ultimate = Cumulative Production + Reserves + Undiscovered

Now comes the interesting part. Many estimates been have made of the world Ultimate for oil, a recent example being the 1995 USGS global survey. The value they published was 2275 Billion Barrels (or Giga - barrels, Gb). These studies are always based on estimates of reserves taken directly from producing countries themselves. Therein lies the problem. Many OPEC countries have been announcing reserve numbers which are frankly very strange. Either their reported reserves remain the same year after year - suggesting that new discoveries exactly match production, or they have suddenly increased their reported reserves by unfeasibly large amounts. This is clearly shown in the following table:

Table 1: Selected Reported Reserves (Gb) with Suspect Increases
YearAbu
Dhabi
DubaiIranIraqKuwaitSaudi
Arabia*
VenezuelaSpurious
Amount
198028.001.4058.00 31.0065.40 163.3517.870
198129.001.4057.50 30.0065.90 165.0017.950
198230.601.2757.00 29.7064.48 164.6020.300
198330.511.4455.3141.00?64.23162.4021.5011.3
198430.401.4451.00 43.0063.90 166.0024.850
198530.501.4448.50 44.5090.00?169.0025.8526.1
198631.001.4047.88 44.1189.77 168.8025.590
198731.001.3548.80 47.1091.92 166.5725.000
198892.21? 4.00?92.85? 100.00?91.92166.9856.30?192.11
198992.214.0092.85100.0091.92169.9758.080
199092.004.0093.00100.0095.00258.00??59.0088.3
TOTALS: Declared Reserves for above Nations (1990) = 701.00 Gb - Spurious Claims = 317.54 Gb
data from Dr. Colin Campbell, in SunWorld, 1995 (click here for references)
In the above table, the red boldface numbers are considered spurious reserve claims. Also curious are the instances of reserves remaining identical over a period of years, despite intensive production. It can be seen that fully 45% of all the above reserve claims are questionable - even neglecting repeatedly unchanged reserve data.

These data are less odd when one realises that OPEC takes into account a country’s reserves when fixing production quotas: the more oil you say you have, the more you’re allowed to sell. Additionally, oil reserves can be used as collateral for loans - an example of this is the $50 Billion loan from the USA to Mexico: in December 1994, the Mexican Peso fell by around 35%. As a result, the Mexican Central Bank's international reserves fell from $29 billion to $5 billion. To stave off a collapse of the Mexican economy, President Clinton signed a $50 billion "Emergency Stabilization Package" loan to the Mexican government on 31 January 1995. The collateral for the loan was Mexico's pledge of revenues from its future petroleum exports.

Another problem with surveys like that of the USGS (from which the US government takes its figures) is that they use very flexible definitions of the different types of oil involved when predicting the amount of oil remaining to be discovered. Briefly, these break down as follows:

  • Conventional Oil (95% of all oil so far produced is conventional)
  • Unconventional Oil
    • Tar Sands
    • Oil Shales
    • Oil not recoverable with today's technology
This distinction is important, because the global economy is based on cheap pumpable petroleum which comes exclusively from conventional oil: there may well be sources of unconventional oil waiting to be found (ie Canada, Antarctica) but not at today’s prices, and not today, either. This counters the argument, often put forward by oil companies, that improvements in technology will prolong the lifetime of our oil resources: the cost of oil produced by these as yet uninvented technologies is likely to be astronomical by today’s standards. It is therefore misleading not to consider these resources as separate from conventional oil.

What is needed therefore, is an estimate of the global ultimate for oil production, which takes into account both ‘political reserves’ and the different kinds of oil that exist.

Reality Dawns:

Such a study has recently been conducted by Dr. C. J. Campbell, on behalf of Petroconsultants of Geneva, and using their data. The Petroconsultants database is the most comprehensive available for data on oil resources outside of continental North America, and is used as a 'bible' by all international oil companies - the information contained in this database is not in the public domain.

The conclusions reached in Dr. Campbell’s study are ominous: He arrives at a figure of 1750 Gb for the global ultimate.

This infers that the midpoint of depletion will occur in 1999.

Figure 2
Graph showing Production (Mbd) against Time (Years), based on Campbell’s data.

The four different lines correspond to different possible scenarios taking place from 1996 onward. It can be seen that whichever scenario actually occurs, the end result is reasonably constant. This is because the Ultimate is a constant value, so that more oil now means less in the future: whilst it may be possible to alter the shape of the curve, one cannot alter the area beneath it. The ‘premature peak’ in the early 1970s corresponds to the oil crisis of 1973.

The Plot Thickens:

This does not mean that the world is running out of oil: it means that we are running out of the cheap pumpable oil that has fueled the economic development of the 20th Century.

The global oil production curve is simply a composite of the contributions of individual nations. However, different countries are in varying stages of production. Some peaked long ago (the USA peaked in 1970 -an event predicted by Dr. Hubbert in 1956), some will peak very soon (the UK in 1999), and some are a long way away from peaking - see graph below. These latter countries will soon find themselves supplying an ever increasing proportion of the world’s oil needs as we pass the global Hubbert Peak.

They are of course the major Middle East producers, the largest of them being Saudi Arabia. Their share of the world oil market will probably exceed 30% in 1999. The last time this happened, in 1973, it allowed them to trigger a world oil crisis. In contrast with 1973, the changes in 1999 will be permanent, as they will be based on resource constraints as opposed to politics.

Time to Depletion Midpoint (Years)

Figure 3
The Above Graph illustrates the time to midpoint for various major oil producing nations.
A negative value means that the midpoint is in the past. The only countries a significant way from their midpoints are the major Middle East producers.

The likelihood of a global crisis similar to that of 1973 is very high. The precise timing is dependent on the durability of Saudi Arabia’s pro-Western stance, and hence on the stability of the current political regime there. This is because Saudi Arabia is pumping more oil than it needs to, in response to Western (mainly US) political pressure (see graph below). In fact, it is highly likely that if Saudi Arabia were to cut its oil output by 20%, it would actually increase revenue from sales, as the resulting supply shortfall would push prices up significantly. This must be very tempting for a country whose debt to income ratio is approximately 2:1. It is highly conceivable that there will be a change of government in Saudi Arabia within 5 years. This will serve to bring forward the onset of a crisis which, as it stands, is coming anyway.

Saudi Exports and US Imports
Figure 4
This graph dramatically illustrates the relationship
between Saudi Exports (black line, black squares) and US Imports (grey line, triangles). (Graph from Campbell, Energy & Exploration, vol.13, no.1, 1995)

The only long term solution is to reduce our dependence on cheap oil from the middle-East. This is an entirely feasible thing to do, however it will require serious concerted action on the part of government, industry, and the financial sector…

updated 2003 December 22