Dec 28, 2011

World Peak Oil

In my last post I wrote a bit about Canadian oil production vs. our proven reserves and gave some projections on how long our reserves might last. Our government says that our oil reserves will last 200 years, however I showed that that is subject to the fragile assumption that there will be no increase in production during that time. The post further shows that over the last 30 years Canadian oil production has averaged about 2.65% growth per year, so using that as a baseline for future oil growth the length of time until Canadian reserves run dry is about 70 years.

This post will blow all that out of the water and show that 70 years is much longer than a more realistic estimate.

I picked up some data from the CIA World Factbook on global oil consumption and proven reserves. From this we can see that world consumption is about 36.75 billion barrels of oil per year, and the total proven reserves is about 1.48 trillion barrels. If we do the simple calculation of just dividing reserves by consumption we get just over 40 years. This means that if oil consumption does not grow any more, we will run out of all the oil we have proven to exist in 2051. This seems like a long time away, however it is within one human lifetime: I will turn 65 in 2051.

That is assuming that there will be no growth in consumption during those 40 years. Looking at this page the average growth rate in oil consumption is 1.18% per year (using a geometric mean since this is a growth rate). If we predict that oil consumption will continue to rise at this rate, we will run out of our reserves a bit sooner: just under 33 years from now, sometime in 2044. This is less than half of our prediction from the Canadian example, and one sixth of the Canadian government's prediction of how long our oil supplies will last.

How is it that the amount of time here is so much smaller than in the Canadian case? Why is it that the world oil reserves will somehow be drained sooner than the Canadian ones, a logical impossibility? The main culprit here is our assumption that Canadian oil production will continue to grow at 2.65% and not something higher. If you look at global oil production by country you can see that four countries (Russia, the USA, China and Iran) all have much higher production than Canada but all have lower proven reserves; they will run out of their oil far sooner than we will. When that happens, consumers around the world who want oil will have to turn to somebody else. Will our 2.65% per year increase in production be able to satisfy those consumers once other sources start running out?

As in the last post, I am stressing that this is not driven by ideology. There is no liberal or conservative bias here, it is just numbers from reasonably good sources and simple, widely known mathematics. You can repeat this analysis yourself and see that you will come to the same numbers. The R code for this one is available here, you'll need the data that I listed above and then convert the CIA files from fixed-width format to CSV format.

Dec 26, 2011

Canadian Peak Oil

One thing that interested me back in my university days was peak oil - the idea that at some point, our ability to produce oil will at some point peak and then decline. It seemed very logical to me that given a finite resource and an ever-increasing demand for it, we will eventually run out. Given the drastic consequences that could happen if we happened to run out of this resource that seems to be used everywhere, it seemed like a good idea to be concerned about this problem and try to figure out how to at least minimize the damage should this disaster scenario come sometime soon.

Today I'm marginally wiser than I was then, so I like to verify myself if these types of things I hear through the media are true or not. I decided to look up some of the details about this whole peak oil thing and wonder, "when would this whole thing start going down?"

I looked on Natural Resources Canada's web page to see some information about oil production. I found this tidbit which says, "Canada's oil reserves are sufficient to meet demand for the next 200 years at current rates of production." Nice! I suppose that means I don't have to care, and let the future generations figure out how they are going to deal with this problem.


Well, maybe not. The first thing we want to do is verify that this number is actually correct. After all, while we know that the government would never lie to us, it's always a smart idea to figure things out for yourself in case somebody over there made a mistake.

To figure out our rate of extraction, this FAQ tells us that Canada currently produces about 2.5 million barrels of crude oil per day (consistent with some data that is mentioned below). To figure out how much oil Canada has proved to have, a 2010 report by the US Energy Information Association tells us that Canada currently has about 178 billion barrels of proven oil reserves that can be extracted, including the oil sands. If you take this number and divide it by the amount of oil we produce each day, we get approximately 195 years of producing oil; the number given by the government website is more or less correct.

Does this mean all is well for the next 200 years? Well, not really. See, the bit that breaks this whole analysis is right in the quote taken from the government's FAQ: "at current rates of production." The 200 year figure assumes that Canadian production of oil will not grow at all over the next 200 years. Intuitively that doesn't seem right, but let's be scientific about this. Maybe it is in fact true that oil production is constant.

Doing a bit of a search with my trusty sidekicks Google and Wikipedia, I found some data here on the website of the Canadian Association of Petroleum Producers that shows the yearly oil production in Canada from 1971 to 2010. Rather than describe it to you, here's a picture:

Canadian Oil Production

If the government's assumption were true, this graph would be a flat line at 2.5. As you can see it is definitely not, in fact it seems to be increasing at a fairly quick rate!

What happens to our numbers if we start increasing oil production each year? Let's find out.

With a linear model, we get that it is increasing by about 0.0336 Mbbl, or 33 000 barrels of oil per year. If this trend continues, it means that the reserves will run out in...194 years. So one year less than if we assumed it would be flat.

Of course, typically in economics it is assumed that economies grow exponentially, not linearly. Since both population and productivity due to technology seem to increase at exponential rates, it a fairly safe assumption to make that production of anything increases at an exponential rate provided nothing stops it.

When we use an exponential model, we can calculate that the oil production is growing at about 1.68% per year. If we extrapolate this out, it turns out our oil supply will run out after...87 years. That's a bit shorter than the last time frame! Even scarier for the people living in that time, we will use up the first half of our current amount in about 60 years, with the second half being used up in a short 27 year time frame! So in short, if production keeps up as it has been, then our grandchildren (assuming you're my age, I'm 25) will see the day when the known oil deposits of Canada will run dry.

In fact, there's another problem with this analysis. The data set we're using includes the 1970's with a massive oil boom followed by a collapse in the mid-70's that lasted until about 1980. If we re-run the exponential model after dropping the 1970's (so we are left with 30 years of data) the rate of increase goes up to 2.56% with an R2 of 0.98 (meaning a near perfect fit). With that rate of increase, our oil will run out in 70 years. This is a lot closer than the 200 years given by NR Canada.

I tried to leave out any speculation from this post and just use numbers and math based on what has been happening and from those numbers, make a simple extrapolation to see what would happen should the existing trends continue. The growth rate of 2.56% per year may change in the future (up or down), which obviously would change the results of this simple analysis.

To add my own little bit of speculation, from what I've seen in the news I think that this rate of production will actually increase over time given the attitudes of the current Canadian government - as more projects such as the Keystone pipeline get completed it will be even cheaper to transport oil into the United States and other countries from Canada. An economics 101 class will tell you that as the efficiency of supply increases, the amount of production will increase: as Canadians get better at shipping the oil to other places, more oil will be produced to ship.

As always with my statistics-oriented posts, you can see the code here. In the code you can see I did a linear model with the 1970's excluded but didn't talk about it here, it's because the exponential model fits both better with the data and better with economic theory.