Have we reached peak oil?

I recently sat down with a group of Morningstar's energy analysts to discuss the idea of "peak oil." In this article, I will define the issue and share the group's insights.

Peak oil defined
At its core, peak oil is the idea that we will reach a point at which the rate of oil production cannot be increased because of geologic limits such as the size of the planet's resource base and the impact of natural decline rates. There are other limits to the rate of production, including above-ground factors such as investment rates and geopolitics, that further constrain production levels. To use an analogy, when thinking about the maximum amount of milkshake one can drink in a certain amount of time, the size of the straw and the ability to suck matters just as much as the amount of liquid in the cup.

The idea behind peak oil is credited to a geoscientist named M. King Hubbert, who worked for Shell back in the 1950s. In 1956 he published a paper that detailed a statistical method he developed suggesting that the rate of fossil-fuel production tends to follow a bell-shaped curve. The idea behind this is that after fossil-fuel reserves are discovered, production begins to increase exponentially until a peak production rate is reached; after that it begins to decline as depletion overcomes new discoveries.

When you look at the history of discoveries, it's pretty clear that we've already found most of the obvious oil fields. In terms of oil reserves, discoveries peaked in the 1960s, and the rate of discovery dropped below our annual consumption in the late 1980s. Today, we're using more oil each year than we find.

2009 was a banner year for oil discoveries, with a lot of headlines being generated by finds in Brazil and the deep waters of the Gulf of Mexico. In fact, we saw discoveries on the order of 10 billion barrels of reserves, the highest rate since 2000 when the giant Kashagan field in Kazakhstan was discovered. However, the world is consuming around 83 million barrels a day, which equates to 31 billion barrels a year. So even in this banner year, we are barely replacing one third of the oil we consume.

Are we running out of oil?
In a word, no. Yet we have essentially found all of the cheap oil. Since Colonel Drake first drilled for oil in Pennsylvania in 1859, the world has used about a trillion barrels of oil. Estimates vary widely, but there are at least another trillion barrels of conventional crude oil reserves and perhaps two or three times that much if you consider unconventional (and higher-cost) sources, such as oil sands and oil shale. We're not going to run out of oil overnight, but it's fair to say that the first trillion barrels we consumed were the cheapest, easiest-to-access reserves.

When you look back at the East Texas oil boom early last century, oil wells were being drilled a few hundred feet deep. In the deserts of Saudi Arabia and Iraq, giant oil fields are so close to the surface that you could practically stick a straw in the ground and strike oil. These big, easy finds were relatively inexpensive to develop.

But check out where we're looking now: The latest Gulf of Mexico discovery, Tiber, is a well drilled to a depth of 35,000 feet and lies beneath 4,000 feet of water. Think about that; the well is a mile deeper than Mount Everest is tall. It will likely take 7–10 years before this discovery produces anything. While this is a significant discovery, it certainly isn't cheap oil.

We have established that cheap oil might be a thing of the past, and it is clear that we are using more oil than we find each year. Yet how does this fit into the notion that oil production is peaking? The key thing to consider is that an oil well's rate of production declines over time.

As oil is pumped from a reservoir, the pressure in the well begins to drop and the rate of flow decreases. This process is called a decline rate. One can drill new wells in a field to balance the impact of declines, but as an oil field is developed and drained from multiple wells, it reaches a point at which the whole field goes into decline. We saw this play out with Alaska's Prudhoe Bay, in the North Sea, and in the Cantarell field in Mexico. Now we can aggregate oil fields and look at production curves for countries in the same way, and we see that 40 of the 54 oil-producing nations are past their peak oil production. In the United States, oil production peaked in 1970 around 9.5 million barrels per day, but today our production is about 5 mb/d.

Let's put oil-field declines in context. World oil production is roughly 83 million barrels per day. Various estimates place the underlying global decline rate somewhere between 4% and 8% per year. That means that each year we have to add about five million barrels of new production to keep production flat. Step five years out, and we have to replace 25 mb/d of production, or about three times Saudi Arabia's current production. That's a lot of new wells that need to be started just to offset declines.

Plus, this does not account for any growth in oil consumption. Absent global recessions, underlying oil demand is increasing by about 1% per year. This means that five years out we'd need another 5 million barrels of oil per day just to keep the current equilibrium. Frankly, we're not certain that we'll be able to reach that level of production.

Have we reached peak oil?
It is hard to tell, and we do not know. No one will know for certain except by looking in the rear-view mirror. A couple of our analysts attended a conference in Denver put on by the Association for the Study of Peak Oil and Gas a few weeks back, and the precise timing of peak oil is of considerable debate. In our minds, the exact timing is less meaningful than the fact that oil production will begin to decline at some point within the next five to 10 years.

One enlightening analysis at the conference was presented by Rembrandt Koppelaar based on tracking announced oil megaprojects and layering anticipated production gains on top of existing world production. His analysis provides a best-case outlook that shows we can bring about 25 mb/d of new production online by 2016, assuming announced projects are completed on time and result in expected new production. His analysis suggests that we will get to roughly 90 mb/d in 2014. Incidentally, this is roughly the level of production an increasing number of oil executives are discussing as a production peak.

The demand side
We've talked a lot about supply issues, but demand is just as critical. Over the past five years we've seen China and other emerging economies bidding barrels away from industrialised countries. In fact, demand from the developed world (defined as the OECD countries) is down by about 4% since 2000, while China's demand is up 60% and India's is up 40%. On a net basis, world demand is up about 8%. In a very real way, the OECD countries have become one of the larger "suppliers" of oil to the market by reducing consumption.

Looking forward, we see this trend continuing, especially if fuel-efficiency measures as well as hybrid and electric vehicles gain traction here. Gasoline consumption in the United States accounts for about 12% of total world demand for oil, and any sizable reduction in gasoline use will free up barrels for the rest of the world. Our efforts to boost efficiency and reduce consumption will certainly affect the supply/demand balance. As Benjamin Franklin might have said, a barrel saved is a barrel earned.

China: The wild card
On the other side of the coin, most of the demand story is China. Formerly an exporter, China became a net importer of oil in 2000. It produces about 4 mb/d but now consumes roughly 8 mb/d. China has been responsible for 4 mb/d of new demand since 2000, about half of incremental demand over that period.

One item worth looking at is the rapid growth in China's car ownership. In March, car sales in China overtook those in the US for the first time, and sales are averaging 1.1 million new units a month. This is roughly twice the level of 2005 car sales. A big driver here was massive government subsidies that make "cash for clunkers" look downright stingy. But the core story of increased affluence, increased urbanisation, and the availability of consumer financing seems to give real legs to Chinese auto demand. Just think, in the auto-loving United States there is a little less than one car per person in the country, but China's ratio is a little over one in 10. It makes sense that both ratios will get closer to one another in the coming years.

Another component here is the fact that China subsidises fuel prices, so Chinese drivers, who pay even less per gallon than drivers in the States, are not very exposed to price increases. If oil prices spike, the price of petrol goes up in the developed world, and there's a demand response (witness 2008). But this impact is muted in China. As long as China can maintain a measure of economic stability, auto demand there is likely to continue its upward trek.

Of course, the big question mark is whether China can maintain its scalding-hot economic growth in the face of much slower growth in the US and Europe, given that China is still export-driven. If China runs out of steam and its GDP drops to, say, 3%–5% annual growth instead of the nearly 9% so far this year, quite a bit of oil demand would come off. China is the wild card.

What happens to oil prices?
Whether or not prices keep escalating to ration tight supply remains the $64 billion question. We do think oil prices are likely to increase as oil supply begins to tighten again, but oil prices are tricky. To some extent, they reflect the state of the dollar. But, perhaps more importantly, high oil prices act as a tax on economies. When oil purchases begin to account for a material level of GDP, say, 4%–6% like we saw in the US in 2008, economies cannot really bear that tax, and demand responds strongly. We don't think that we're likely to see oil shooting past $200 a barrel. Instead, we tend to think that high prices will cure high prices and lead to reduced demand. Some analyses we've seen suggest that $100 oil is enough to trigger another recession in the US. So high prices are likely to throw countries back into recession and reduce demand that way, which will result in a lower oil price.

The trick is this: Companies need oil to be above $60 or so to bring new production online. So we're on a narrow plank between the price required to bring on new production and the price that throws us back into a recession. And as we continue to push the frontier and supply tightens, the price to bring new production online begins to increase and the plank narrows. What do we do when producers require $80 oil to add new production, but oil prices that high keep us at recession's edge? We're certainly in a box that will be painful, but not impossible, to get out of.