Will the decline in world oil supply be fast or slow?

I wrote a post last week called Steep oil decline or slow oil decline? Since writing it, I had some additional thoughts on the subject, on reasons to expect a steep decline rather than a slow decline.

Furthermore, my article What’s behind United States budget problems? got me to thinking about the reasons for declining employment. In this post, I also try to explain the connection between declining EROEI , declining demand for oil, and lower employment. I explain why free trade with China and India tends to make the employment problem worse and increase global CO2 emissions. It also increases the chance of collapse in the developed world.

A reader wrote, asking the following question:

Dear Gail,

I have been reading quite a bit about peak oil recently. I get the impression (not based on data) that at some point there will be a quite steep decline in oil production/supply, and therefore we will see dramatic changes in how the world runs. However, when I look at oil depletion rates and oil production declines based on the Hubbert Curve, it seems to suggest a rather smooth decline.

How is that some people expect a serious energy crunch in about two or three years, then?

Many thanks!


My (longer) answer is as follows:

Dear Sam:

It seems to me that

(1) A slow decline assumes that the only issue is geological decline in oil supply, and the economy and everything else can go on as usual. Technological advances and switches to alternatives might also be expected to help keep supply up.

(2) A fast decline can be expected if one or more adverse factors make oil supply decline faster than geological factors would suggest. These might include:

(a) Liebig’s Law of the Minimum – some necessary element for production, such as political stability, or adequate food for the population, or adequate financial stability, is missing or

(b) Declining Energy Return on Energy Invested (EROEI) interferes with the functioning of society, so the society generates too little net energy, and economic problems ensue, or

(c) Oil becomes so high priced that there is little demand for it. This would quite likely be related to declining EROEI.

My view is that some version of the faster decline scenario is likely, because we will hit limits that interfere with oil production or oil demand.

Let me explain my reasoning.

Declining EROEI

EROEI means Energy Returned on Energy Invested. It can be defined as the ratio of the amount of usable energy acquired from a particular energy resource to the amount of energy expended to obtain that energy resource. Wikipedia says,

When the EROEI of a resource is equal to or lower than 1, that energy source becomes an “energy sink,” and can no longer be used as a primary source of energy.

The situation is really worse than Wikipedia suggests. An economy needs a certain level of energy just to keep its infrastructure (roads, bridges, schools, medical system, etc.) repaired and working, and citizens educated. So energy resources, to really be useful, need an EROEI significantly higher than 1 to maintain the system at its current level of functioning.

How much higher than 1.0 the EROEI needs to be on average will depend on the economy. An economy such as that of China, with relatively fewer paved roads and less expensive schools and healthcare system can probably get along with a much average lower EROEI (perhaps 4.0?) than an economy like the United States (perhaps 8.0), because of lesser infrastructure demands.

If the average EROEI available to society is falling because oil is becoming more and more difficult to extract, an economy with a high standard of living such as the US would seem likely to be affected before an economy with a lower standard of living, such as China or India or Bangladesh, because of the higher EROEI needs of the more extensive infrastructure. Ultimately, though, the world is one economy, so problems in one country are likely to affect the economies of other countries as well.

There a couple of issues related to declining EROEI:

1. High cost to extract. Sources of oil or natural gas or coal that are difficult (high cost) to extract tend to be lower in EROEI than sources that are low-cost to extract. So high cost of extraction tends to be a marker for low EROEI. We are increasingly running into this issue, for both oil and natural gas.

2. Declining Net Energy. EROEI is closely related to “Net Energy,” which is the amount of usable energy that is left after deducting the energy that it takes to make energy. When net energy decreases, we have less energy to run society, making it difficult to do things like maintain bridges and roads, and fund schools.

So high cost of oil extraction, low net energy, and low EROEI are all very closely related.

What did M. King Hubbert Say?

M. King Hubbert in various papers such as these (195619621976) talked about a world in which other fuels took over, long before fossil fuels encountered problems with short supply.

Figure 1. Figure from Hubbert 1956 paper, Nuclear Energy and the Fossil Fuels

In such a world, there would be plenty of net energy from alternative fuels to run society. Because of this, even if fossil fuels ran low, it would be easy to maintain the economy’s infrastructure, without disruption. In Hubbert’s 1962 paper, Energy Resources – A Report to the Committee on Natural Resources, Hubbert writes about the possibility of having so much cheap energy that it would be possible to essentially reverse combustion–combine lots of energy, plus carbon dioxide and water, to produce new types of fuel plus water. If we could do this, we could solve many of the world’s problems–fix our high CO2 levels, produce lots of fuel for our current vehicles, and even desalinate water, without fossil fuels.

He also showed this figure in his 1956 paper:

Figure 2. Figure from Hubbert’s 1956 paper, Nuclear Energy and the Fossil Fuels.

In this figure, most of the additional energy comes from nuclear energy, while a smaller amount comes from “solar” energy. By solar energy, Hubbert would seem to mean solar, wind, tidal, wood, biofuels, and other energy we get on a day-to-day basis, indirectly from the sun. His figure seems to suggest that solar energy would basically act as a fossil fuel extender, and would not last beyond the time fossil fuels last. The primary long-term source of energy would be nuclear.

Figure 3. Hubbert’s view of world oil production from his 1956 paper, Nuclear energy and the fossil fuels.

In such a world, applying Hubbert’s Curve to world oil supply would make perfect sense, because there would be plenty of other energy, to provide the energy needed to keep up the infrastructure needed to main extraction of oil, gas, and other fuels as long as they were available. Even liquid fuels and pollution wouldn’t be a problem, if they could be manufactured synthetically. The carrying capacity of the world for food would eventually be a factor, but in one scenario in his 1976 paper, he shows the possibility of world population eventually reaching 15 billion people, thanks to the availability of other fuels.

Another Approach to Forecasting Future Oil Supply: Limits to Growth Type Modeling

Another approach estimating the shape of the decline curve is by applying modeling techniques, such as used in the 1972 book Limits to Growth by Donella Meadows et al. The factors considered in this model were population, food per capita, industrial output, pollution, and resources. Resources were modeled in total, not oil separately from other types of resources. There were 24 scenarios run. The base scenario suggested that the world would start hitting resource limits about now (plus or minus 10 or 20 years). There have been several analyses regarding how this model is faring, and the conclusion seems to be that it is more or less on track. This is a link to such an analysis by Charles Hall and John Day.

With this type of model, according to Limits to Growth (p. 142), “The basic mode of the world system is exponential growth of population and capital, followed by collapse.” This type of decline would seem to be substantially faster than the decline predicted by the Hubbert Curve.

One thing I notice about the Limits to Growth model is that it leaves out our debt-based financial system. Since so much capital is borrowed in today’s world, it seems like including such a variable would tend to make the system even more “brittle”, and perhaps move up the date when collapse occurs.

Also, the Limits to Growth model is for the world as a whole, rather than for different parts of the world. Different areas of the world can be expected to be affected differently, as oil gets in shorter supply. The effect of this would seem to be to push economies which have a higher need for oil (illustrated above with my estimate that he US requires a EROEI of 8.0 on energy resources) down toward economies that use smaller amounts of oil (illustrated by my rough guess that perhaps China could get by with an EROEI of 4.0), especially if they trade with each other. I explain how I see this happening in a later section of this post.

Demand for Oil (or other Fossil Fuels)

Even if there is plenty of high-priced oil extracted from the ground, if potential buyers cannot afford it, there can be a problem, leading to a decline in oil production. Demand can be thought of as the willingness and ability to purchase oil products. Many people would like to have gasoline for their cars, but if they are unemployed, or have a part-time minimum wage job, they are likely not to have enough money to buy very much.

Over the long term, declining demand can be expected because of declining EROEI, as illustrated by Prof. Charles Hall’s “Cheese Slicer” model.

Figure 4. Professor Charles Hall’s cheese slicer model of the economy, reflecting the energy needed to make energy, and other aspects of the economy at 1970

Figure 5. Professor Charles Hall’s cheese slicer model of the economy, reflecting the energy needed to make energy, and other aspects of the economy at 2030.

Declining demand, and ultimately lack of sufficient demand to support supply, is related to the much larger size of the big black “energy needed to create energy” arrow as resources become more and more difficult to extract, and the much smaller size of the red discretionary spending arrows. When the discretionary spending arrows are small, people can’t afford the oil that is produced.

Lack of Demand Can Be Expected to Affect the More-Developed World before the Less-Developed World

Let me explain one way I see lack of demand for oil arising in the developed world today. This is related to the tendency of economies with high required EROEI to maintain infrastructure to be the first economies to be affected by declining EROEI, and by the tendency of free trade to lead to equalization among economies.

Figure 6. US energy consumption, from Energy Export Data Browser

US energy consumption in general, and oil consumption in particular, has been relatively flat in the 2000-2009 period, and declining at the end of that period, indicating low demand. Prior to this period, it was rising.

More or less the reverse has happened in China and India. Growth in oil use and energy products in general was moderate prior to 2000, but increased rapidly after 2000.

Figure 7. China’s energy consumption, from Energy Export Data Browser

Figure 8. India’s energy consumption, from Energy Export Data Browser

When we look at the percentage of the US population that is employed (Figure 9), it has been decreasing since 2000, so there are fewer people earning wages, and thus able to buy oil and other products. Prior to 2000, the percentage of the US population working was increasing.

Figure 9. US Bureau of Labor Statistics employer non-farm employee counts divided by US Census Bureau resident population estimates

In fact, over time, in the US, there is a high correlation between number of people employed and amount of oil consumed.

Figure 10. Number of jobs from US Bureau of Labor Statistics non-farm employer; oil consumption is “Product Supplied” from US Energy Information Administration

This high correlation is not surprising for two reasons: (1) jobs very often involve often use oil in producing or shipping goods, and because (2) people who are earning a salary can afford to buy goods and services that use oil.

If we think about it, businesses employing people in China and India have three cost advantages over businesses employing people in the US:

1. People in China and India earn less, in large part because their life styles use less oil. As the price of oil has rises, a person would expect this difference to become greater, if salaries of US earners are raised over time, to reflect the higher cost of oil, as it rises. If the living standards in China increase, the salary differential could decline, but still might be very high in dollar terms.

2. The cost of electricity used in manufacturing in China and India is cheaper, because it is generally coal-based. The cost of electricity from coal is quite likely even cheaper than electricity from coal from the United States, because these countries are more likely to have poor pollution controls, and because the coal is extracted using cheap labor. The difference in the cost of electricity can be expected to become greater, to the extent the US imposes stricter pollution regulations, or switches to higher priced alternative power (say, offshore wind), or imposes a carbon tax.

3. Taxes and employee benefits are likely to be lower (in absolute dollars, but perhaps as a percentage as well) in China or India, because infrastructure is less complex, and because there is less in the way benefits comparable to Social Security, Medicare, etc. (This is related to the lower EROEI required to maintain the infrastructure in these countries.)

With these advantages, as trade restrictions are eased and more “free” trade of services is enabled through the Internet, I would expect an increasing number of jobs to move overseas, and more goods and services to be imported. Salaries will also tend to stay lower in the US, especially for jobs associated to goods and services that can be produced more cheaply in China or India.

With these lower salaries in the US, demand for oil in the US will tend to be lower, because people who are paid less (or out of work) will not be able to afford high-priced oil for vacations and other optional purchases. As more US jobs move overseas, unemployment and recession can be expected to increasingly become problems. Furthermore, it will become difficult to collect enough taxes from the lower number of employed people to pay enough taxes to keep the system operating. I write about this in What’s Behind the US’ Budget Problems?

One thing that happens, too, with this arrangement is that world’s coal use has risen.

Figure 11. World energy consumption, from Energy Export Data Browser

I wonder if all of the emphasis on CO2 reduction has not exacerbated the problem. Countries that reduce their own coal use and instead rely more on exports can feel virtuous, but they also set the stage for negative impacts. By using less coal, these countries leave more coal for lesser developed countries to import. These lesser developed countries probably burn it less safely (for example, with less mercury controls) and compete with them for jobs. The developed countries can be expected to have more and more budget problems, as their tax bases erode, and the number of unemployed rises.

When new electricity generation is planned in the United States, the usual practice is to compare expected costs with other types of new electricity generation that might be possible in the United States. It seems to me that this practice does not show the full picture. Goods and services produced in the United States will have to compete with goods and services produced around the world. Some of the electricity used will be from nuclear plants that have long been paid off; some will be from coal production; and a little will be from high priced new types of electricity production. As long as there are no tariffs or other trade restrictions, higher-priced US electricity will tend to hinder exports and help imports. I would vote for trade restrictions.


The downslope of oil production can be expected to reflect a combination of different impacts. Unless technology improvements truly have a huge impact, it would seem to me that the overall direction of the downslope is likely to be faster than Hubbert’s Curve would predict.

Thanks for writing!

Best Regards,

Gail Tverberg (also known as Gail the Actuary)

This entry was posted in Financial Implications, Oil and Its Future and tagged , , , , by Gail Tverberg. Bookmark the permalink.

About Gail Tverberg

My name is Gail Tverberg. I am an actuary interested in finite world issues - oil depletion, natural gas depletion, water shortages, and climate change. Oil limits look very different from what most expect, with high prices leading to recession, and low prices leading to financial problems for oil producers and for oil exporting countries. We are really dealing with a physics problem that affects many parts of the economy at once, including wages and the financial system. I try to look at the overall problem.

37 thoughts on “Will the decline in world oil supply be fast or slow?

  1. >>
    I tend to think the ruling goverments of the US and China ( and also Russia) will be too circumspect to provoke each other. There is little to be gained and too much to be lost by direct nuclear exchanges between superpowers.

    But this is Cold War thinking. Not post Peak world thinking.

    This is not to say you’re wrong. Nobody wants to go to war for the sake of going to war. The maneuvers for advantage in the Cold War were ideological. Only in the end game were they economic, as the Soviets approached surrender. There is a phase of the decline where I can envision your point being in play, but it seems brief.

    Establishing the ground rules is the initial truth. People will fight to survive. They will kill to survive. They will have remorse about it that night, and the next day they will go out and kill more. If the US maneuvers to retain its 4/24 configuration, someone out there will be starving early, and they will fight to survive.

    It is at this point that your point has its phase. The US government will have to make a cold calculation that it must reduce its voters quality of life to . . . not prevent that starvation out there, but to prevent the starving people from waging war. Compassion won’t be the motivator. We let millions starve in Africa without any national gut-wrenching guilt. We won’t be sacrificing oil to prevent starvation. But we may sacrifice it to prevent suicide bombs from attacking oil production facilities.

    That will be a delicate debate, though. If it turns into . . . we must reduce our quality of life because we’ve been threatened with war, then it’s just Cold War tension again and no oil will be sent there.

    The Russians are sitting pretty, unless their production avalanches for lack of investment. They are holding their numbers up with 6000 wells drilled per year. The Western companies trying to get a piece of that action (BP and RDS) might not like the $5 or $10/barrel they get to keep for that enormous investment. Rosneft can’t do the work themselves or they wouldn’t be hunting partners. Regardless, if the western OCs take their marbles and go home, Russia production will crash.

    No, it’s all about China, and then India on the back burner. It’s consumption. We all know nothing can be done about supply. It is consumption that has to crash. Quickly.

    There’s only one way for that to happen, and it’s not windmills.

  2. It seems to me that in the United States, the peak was fairly sharp because it was unexpected (except by Hubbert and a few others), and the rate of decline was fairly slow because over the decades following the peak, a number of techniques for increasing ultimate recovery and maintaining a higher rate of production were invented and applied. For increasing ultimate recovery; water flooding, maintaining the gas cap, steam flooding, adding carbon dioxide to lower viscosity, and fracking to increase porosity. For maintaining flow until near the end of a given well: mainly horizontal drilling guided by improved seismic reading of the reservoir, especially bottle-brush wells.

    Because these techniques have been invented before the peak in Middle Eastern and some other oil fields, and the managers of the those oil fields are aware of the facts of depletion, it seems reasonable that the overall effects would be different – flattening of the peak, followed by rapid decline when all known methods for improving recovery are exhausted. This is consistent with rapid decline in the North Sea, in Cantarell, and other examples.

    That’s the general picture that I have. It is muddied, of course, by all the “above-ground” factors that make actual outcomes erratic – infrastructure problems, political problems, discovery of new oil fields, short-sighted and incompetent management of oil fields in general, and so on. As soon as we see rapid decline in the Middle East, there will be plenty of infrastructure and political problems to intensify the rate of decline and mask the cause.

    That’s the simple outline I have developed over five years of paying attention to The Oil Drum and otherwise trying to understand the issues.

    Art Myatt

    • Maybe what we have a situation where supply is really impacted two ways –declining supply, as geological limits are hit; and messed up demand, because of all the financial problems of buyers, and food riots, and international trade problem. It is pretty clear we are hitting a crisis soon, but it is really from multiple effects.

  3. I think governments will do desperate things post peak energy. I think they will be less military and more things like building lots of nuclear power plants.

Comments are closed.