The annual production of a resource such as crude oil depends on several factors including

• the amount of oil still in the ground,
• the cost of extracting it, and
• the availability and price of global alternatives.

In the simplest case, where alternatives are considerably more costly to produce, annual production would be expected to rise over time as the large oil fields are discovered and the technology of consumption (such as motor vehicles) are developed and deployed.

Eventually, older fields reach their peak and production begins its inevitable decline. During the decline, technologies such as water flooding and gas injection may be introduced to slow the rate of depletion of older fields.

Meanwhile, exploration continues for new oil fields. Eventually, though, new fields become too small and economically not worth developing.

Overall production continues to fall, and substitutes less costly than crude oil are developed and introduced. In the case of the coterminous 48 states (the "lower-48" states in oil-industry language), imported oil turned out to be the cheapest and most convenient replacement source when production began its decline in 1970.

Production of a non-renewable resource

shows schematically in the simplest of cases how annual production might look. Note that the area under the curve -- i.e., the sum of annual production for all years -- is equal to the total amount of oil that will be produced over all time, the Estimated Ultimately Recoverable (EUR) oil.

U.S. lower-48 crude oil production

and Alaska crude oil production show the actual production data for the lower-48 states and Alaska. The smooth curves are "best" fits to the annual production data using simple bell-shaped curves.

If annual oil production follows a simple curve like that shown in Production of a non-renewable resource

, then cumulative production -- the total amount of oil produced to date -- should follow a simple "S" shaped curve (like a logistic curve) that would eventually level off at the value of the ultimately recoverable oil.

As Cumulative production of crude oil in the lower 48 states

and Cumulative production of crude oil in Alaska show, the data for cumulative consumption do indeed closely follow such simple curves.

• Ultimately, these curves suggest, recoverable crude oil in the lower-48 will total about 190 billion barrels (bbl).



• Of this amount, 166 billion (87 percent of the total) had been produced by December 1999.



• For Alaskan oil fields, about 14.5 billion barrels will ultimately be produced, of which 13.8 billion (95 percent of the total) had been produced by the end of 1999.

The conceptual basis for the simple model of oil production outlined here was reviewed in some detail by Cleveland and Kaufmann, both researchers at Boston University's Center for Energy and Environmental Studies.[1] They noted that:

• the model does not explicitly take economic factors into account but that the overall pattern (in U.S. lower-48 crude oil production

) of annual production was correct.

deviations between predicted and actual production could, they assert, be explained largely by considering political events and changes in world oil prices.

the bell-shaped model embodies the long-run effects of technical change and resource depletion and that this empirical approach has proven more accurate in its predictions than many econometric and geological models.

Cleveland and Kaufmann concluded that there is no a priori reason why oil production should follow such a smooth, symmetric bell-shaped curve. And, indeed, the global oil production data shown in World crude oil production

do not.

What they do reflect is the political and economic turmoil of the past three decades. In particular, the price shocks of 1973 and 1979 contributed to economic disruptions and the substitution for oil in some countries of natural gas, coal, nuclear power, and increased efficiency.

To gain some insight into roughly how much oil remains to be produced in the world, and how global production might evolve over the next few decades, we examine various estimates of global ultimately recoverable oil.

Notes

1. Cutler J. Cleveland and Robert K. Kaufmann, "Forecasting Ultimate Oil Recovery and Its Rate of Production: Incorporating Economic Forces into the Models of M. King Hubbert," The Energy Journal, 1991, vol. 121, no. 2: 17-46.