The difficulty of identifying *the* peak of oil production is that there
are many different reasons we might reach *a* peak. Most of those
reasons don't definitively say there won't be a new peak a few years or
a few decades or more later. However some reasons are more interesting
and more oil-related than others, and are therefore more likely to cause
long lasting peaks.
A useful comparison might be copper or tin. I wish I could find some
data to back this up, but presumably production of both peaked shortly
before the bronze age collapse and those peaks were probably not
surpassed for many hundreds of years afterwards. With the benefit of
hindsight this was a very significant peak, but the bronze age itself
lasted thousands of years of turbulent human history and would have
experienced many localised copper/tin mining peaks too as civilisations
and trade routes rose and fell. Although it would still have been very
hard to spot at the time, the defining clue that the last peak was
sustainable was the shift to iron use.
Back to oil, the current biggest driver of increasing use & production
seems to be population growth. Oil is famously price inelastic.
Collectively we go to great lengths to get our daily fix but we have
done for a long time now. Per-capita oil production has been remarkably
stable for several decades (masking some regional variation) since a
probable peak around the 70s oil crisis. World population and oil
production rise together and so they may well plateau, or fall together.
I think in trying to predict peak oil production there's a real danger
we just predict a flu pandemic or significant nuclear exchange that
causes a >5% population drop. Followed by a new peak oil after
Conversely, if attempts to combat climate change bring about a global
per-capita fall in oil consumption & production of 10% over 6 years that
might be a kind of bronze-to-iron signal that's very interesting and
predictive of a long-lasting peak, but with natural population growth of
~1% a year it still wouldn't represent a 5% drop in total oil production.
Even if we're interested in testing a supply-driven peak oil (as opposed
to a politically lead demand-driven drop) population is still a problem
because more bodies pulling smaller quantities of more expensive oil out
of the ground could mask an otherwise clear 'finite oil' signal. (Oil
prices would be more interesting here, if it weren't for that price
inelasticity & associated volatility.)
So this is really a long-winded way of saying wouldn't it be better to
factor out world population?
Another half-idea is to look at oil as a percentage of total world
energy use. The 70s oil peak appears to have coincided with a temporary
energy-use peak. In other words oil wasn't replaced in the energy mix,
we just used less oil and energy temporarily, then transitioned to
making more efficient use of it. Those efficiency gains were never
reversed but the event didn't lead to a sustained drop in oil use
because it wasn't really replaced by anything better. More recently oil
as a share of total energy has decreased significantly, presumably
caused by the relatively cheap cost of other fossil fuels, and with
increasing renewable energy production. Looking at how this trend plays
out could be interesting, but this is probably another claim entirely.
It would rest on finding solid figures for all kinds of energy and
pinning down what type of energy measure really matters, which might be