Its fairly common these days to hear references to "peak oil theory" as a reason why the world is headed towards doom in the near future as we run out of oil and energy prices skyrocket. At the very least, I'm continuously bombarded by books about this when I browse the current-affairs section of the local bookstore. As far as I can tell, theres not that many scientists who literally believe in this; theres a few geologists, an economist here and there, and many environmental activists who promote "peak oil" in order to get more government investments into solar and wind power.

There's various debunkings of peak oil theory available online (see links at the end), but I wasn't completely satisfied by them. So here is my own summary of the basic arguments involved (and their flaws).

First, here is the peak oil argument, at least as I've managed to understand it.
It begins with the uncontroversial fact that petroleum production at many particular oil fields follows a bell-shaped peak. Here for example is the petroleum production of the North Sea:





Moreover, this pattern often holds up when we look at groups of oil fields. For example, if we look at all the oil fields in the US:



The peak oil people then assert that the world's oil production must be following a similar curve. They try to guess how much oil we have left by fitting bell-curves to the currently available data.

This last part is, in my opinion (and in the opinion of many experts), very problematic. Even if we supposed that oil production in any particular source follows a bell curve, it doesn't follow that the world's production will follow a bell curve. The reason people stop extracting oil from any particular place is that it gets more expensive than other places, which are continually being discovered and remapped. Moreover, technology changes all the time, thereby changing which places are the cheapest. As Ismael Hossein-Zadeh, an economist at Drake University puts it,

One of the major defects of Peak Oil is its facile extrapolation or transition from micro to macro level, that is, an unwarranted generalization or extention of what is true in the case of an existing oil well or oil field to the entire world oil production. It is true that every operating or producing oil well or field increases in production rate until it reaches a maximum or peak flow rate, after which the rate of production enters a terminal decline. It does not follow, however, that global world oil production as a whole must soon reach a maximum and begin to run out afterward...

In fact, even as individual oil fields have "peaked" world's oil reserves have been increasing. As this Reason article explains,

...petroleum optimists, such as the analysts at the [US Geological Survey] ... point out that reserve growth and new discoveries have been outpacing oil consumption. (Reserve growth is the increase in production in already discovered and developed fields.) From 1995 and 2003 the world consumed 236 billion barrels of oil. It also saw reserve growth of 175 billion barrels, combined with 138 billion barrels from new discoveries, added a total of 313 billion barrels to the world’s proven oil reserves.

To summarize, we keep discovering new oil, as well as new methods for extracting it, which is why individual oil fields tend to peak and decline, even as global production shows no sign of slowing down and global reserves increase.

Let's go back to the data. Here is the graph of global oil production:



So far, it does not look like the left-half of the bell curve. In fact, the main thought I have from staring at it is that it sure looks like oil production mirrors economic growth. The 50s, 60s, and early 70s were a time of rapid economic growth for the world and oil production expanded fast. Then came the slump of the late 70s and early 80s, followed by slower economic growth ever since. Note the lack of growth in the last year or so corresponding to the current economic crisis.

In fact, peak oilers have a pretty bad record in thinking each hill in this graph represents the permanent peak of world oil production. Just to point out a couple of of examples from especially prominent peak-oilers: Colin Campbell predicted that oil peaked in 1989 and Ken Duffeyes predicted that oil peaked in 2000.

The bottom line:

I believe the peak-oil arguments are pretty silly. Let's not forget, of course, that the amount of oil in the ground is finite, so its bound to run out eventually. However, if your goal is to understand when this will happen, trying to fit bell curves to current oil production is a waste of time.

Finally, if you are interested in more, here are some links to others who have made these and other criticisms. The Reason article mentioned above is a good place to start. This article from the economist focuses its attention on the massive alternative energy sources (tar sands, gtl, shale oil) which will most likely overtake conventional crude oil in the next few decades, and which are either neglected or vastly underestimated in most peak oil arguments. Finally, there is this economics-focused takedown from Ismael Hossein-Zadeh which I quoted from earlier.