Friday, February 06, 2004

Is A Peak in Global Oil Production Looming?
I've alluded several times to the subject of oil depletion, decline curves, and the controversy over a possibly imminent peak in global oil production that would presage an eventual decline. While I don't have the geology background to weigh in on the Hubbert/Deffeyes/Campbell argument, my own experience in the planning, economics, and logistics side of the industry suggests a useful qualitative way to look at the issue. In this view, serious consequences don't depend on a geological limitation on the amount of oil that can be extracted from the ground, but rather from a set of practical and much more mundane constraints.

Oil demand has increased in the last decade and looks set to continue doing so, not incredibly fast, but steadily, about 1-2%/year. Efficiencies in the established economies are likely to be offset by new demand, as countries like China and India develop. As a result, even if global oil production doesn't hit a theoretical peak and start to decline, it could well reach a point at which it can't keep up with the growth in demand. Depletion plays a significant role in this story, but in a subtler way.

Start with current oil production--and demand--of roughly 80 million barrels/day (which I will abbreviate as MBD). Experts see this figure growing steadily towards 100 MBD in the next decade or two. Now add up all the places that can bring on new production to deliver the required incremental 20 MBD. Clearly this would have to include countries such as Saudi Arabia, Iraq, Russia, Nigeria, Angola, Venezuela, and the Caspian Sea region. Now factor in the need to find new production to replace the amount by which current production will have slowed down by then, which could be as much as 30-40 MBD.

In other words, in order to reach 100 MBD in 20 years, we will have to find NEW production that amounts to more than half of current production. That is an enormous challenge, and I'm skeptical it can be done in the real world, when you factor in the kind of practical constraints the industry faces, such as:
- Disputes and delays affecting pipeline routes, e.g. for Caspian oil
- Government policies in the Middle East and Mexico that keep out international oil companies and their capital
- Legal and human rights challenges to new oil and gas projects all over the world
- Environmental and land use restrictions that put some reserves out of bounds (e.g. offshore Florida and the Arctic National Wildlife Refuge
- Capital markets that see opportunities and returns in other industries as more attractive than those in the oil industry

Oil economics are also peculiarly shortsighted, for such a long-term business. The commodities markets focus on today's supply and demand picture, rarely looking more than a couple of years out. And the farther you go out into the future, the thinner these markets become. If there is a production peak or supply/demand gap sitting out there, whether due to geology or geopolitics, and if it's not blindingly obvious until you have actually hit it, than the economic signals from the oil markets telling you to find an alternative are going to come pretty late in the game, too late to provide time for a transition to something else.

Most of us who have looked seriously at a potential hydrogen economy believe it will take a minimum of 15-20 years, and possibly as long as 30 years, to make that kind of transition, because of fleet turnover issues and the high costs and low returns associated with much of the required infrastructure. Even government edict probably couldn't force it through in less than a decade, under nearly wartime discipline.

So if there's a peak in oil production or a serious supply shortfall in the years ahead, then the market will go through a true discontinuity when it occurs. That is, prices on the far side of such an event will be much higher than prices on the near side, probably permanently. These kinds of discontinuities change the world, just as the perceived oil crisis of the 1970s would have, had it lasted more than a few years.