Bill Totten's Weblog

Saturday, July 23, 2005


by David M Delaney

Price does not create oil (April 02 2005)

Don't make the economists' mistake of thinking of peak oil primarily in terms of the price of oil. They refuse to think of it except in terms of price because they believe as a matter of professional ideology that the supply of individual resources cannot limit economic activity. They believe in the fungibility of everything. The classic economist's comment is "Peak oil does not matter because energy is much smaller proportion of GDP than it used to be". The innocents who say this do not understand that NOTHING happens without energy, and that decreasing energy supplies mean a contracting economy no matter what fraction energy, as such, is of the economy.

As to the rate of the decline of oil production: ASPO projects a relatively constant rate (absolute rate, not percentage) decline of "all liquids", which includes "heavy oil" and the tar sands and natural gas liquids, of about 1.4 million barrels per day per year, or about 100 gigawatts per year thermal. They project this relatively constant rate of decline to last for about forty years before it starts to tail off into ordinary exponential decline. This may turn out to be an extremely conservative estimate for several reasons. We will find out very quickly by watching the Cantarell field in Mexico. This field's production is exceeded in the whole world only by the Ghawar field of Saudi Arabia. It has now started into decline. We will find out in, say, three years the implications of the advanced recovery methods to which this field (and all old giants) have been subjected. Many experts believe that these methods have ALREADY postponed the peak of most of the world's old giant and super giant fields that produce half of all the oil, and that, as a consequence, we are due for a rapid collapse of each of each of these old giants once its eventual decline starts. This is an apocalyptic projection. it could mean a decline of five percent per year or more within ten years.

{ASPO is the Association for the Study of Peak Oil and Gas}

But suppose ASPO's moderate decline projection (about two percent per year) is valid. This is a deficit of about four percent per year between supply and the requirements of a growing economy. This is hardly less apocalyptic when you consider its economic effects. There is no way for humans to escape a contraction of everything they do in the face of such a contraction of the oil supply. The question then becomes whether our global culture can keep working when its growth ethos is forcibly removed. In principle, adaptation to a slow decline of economic activity seems possible, at least on the basis of purely physical questions. But when you look at how economic growth is built into the foundation of every large human institution, you have to think that we are in for much bigger problems than mere adaptation to new physical circumstances. It seems that the foundation institutions of our global cultural may stop working suddenly, in the space of a year or two, as the motivations that sustain them are seen to have disappeared - banking, capitalism, social peace and democracy based on the expectation of bigger pies.

Lots and Lots of Square Meters of Solar Panels (April 2005)

This is for those who enjoy playing around with numbers. David calculates how many square kilometres of photovoltaic panels would be required to replace the energy loss that a two percent annual drop in oil production would generate.

To keep the sum of wind electricity power, solar electricity power, and oil power constant at its 2017 value (somewhat less than today) during the first forty years of decline of oil production would require 1000 square kilometers of solar panels AND 87,000 very big (three megawatt) wind turbines to be installed in every year of the decline.


In this calculation I use the form of scientific notation for numbers found in several well known computer programming languages, in which "xey" means "x times 10 raised to the power y". For example 1e9 = 1 times 10 to the 9th power = 1 billion.

ASPO says we can expect a decline of world oil production (all liquids) of 2% per year in 2017 when production is passing downward through 25 billion barrels per year (25e9 bbl/yr), and that the volume rate of decline (not the percentage rate) of the maximum possible production will be fairly constant for decades. (This assumes, of course, that we will maintain sufficient social cohesion to be able to maintain maximum possible production.)

This rate of production decline is 0.02 x 25e9 = 0.5e9 = 0.5 billion bbl/yr/yr. (1.4 million bbl/day/yr) (For data I measured the graph in the ASPO Newsletter.)

This rate of oil production decline is equivalent to a rate of energy production decline of 0.5e9 bbl/yr/yr x 6e9 joule/bbl = 3e18 joule/yr/yr.

Expressing this rate of decline of energy production in watts/yr: 3e18 joule/yr/yr /(365 x 24 x 60 x 60)= 1e11 joule/s/yr = 100 gigawatt/yr.

To maintain a flat energy availability to humans (no decline AND no growth) would require, for example, the commissioning of 100 one-gigawatt nuclear plants per year, or 400 gigawatts per year (400,000 megawatts per year) of solar electric panel nameplate capacity or wind turbine nameplate capacity. This does not include the effects of the decline of natural gas production. The US, as the user of 25% of the world's oil production, would, sooner or later, need to build 25% of the replacements.)

ASPO projects that oil production will be down to 25e9 bbl/yr (the data point used for the calculations above) in 2017, and that production will then have been declining at about 0.5 billion bbl/yr/yr for about ten years, and will continue to decline at about the same rate for another thirty years.

So, how big is the required ongoing construction project if solar and wind electricity are used to replace the power loss from the decline of oil?

An article at

lists the following countries three and their total installed solar electric capacity (nameplate) as being the countries with the greatest installed capacity.

Japan 1.13 gigawatts
Germany 0.70 gigawatts
USA 0.37 gigawatts

Total 2.2 gigawatts

Let's guess that the rest of the world has almost as much again, for an approximate world total of four gigawatta installed solar electric nameplate capacity.

This means that to offset the decline of oil with solar electric power would require the installation EACH YEAR FOR FORTY YEARS of 100 times as much solar electric capacity as has ever been installed to date. (Assumes 100 gigawatts per year required, capacity factor of 0.25, giving a required nameplate solar panel capacity of 400 gigawatts per year).

An article at

lists the global installed wind turbine capacity (nameplate) by country, giving a total of 7.5 gigawatts.

Again assuming a capacity factor of 0.25 and therefore a requirement of 400 gigawatts wind turbine nameplate capacity per year. To offset the decline of oil with no growth of energy use would require the installation EACH YEAR FOR FORTY YEARS of 400/7.5 = 53 times as many wind turbines as have ever been installed to date.

Let's assume a combination of the two kinds of renewable energy, solar electric and wind, installed in the same proportion as to date, total 11.5 gigawatts, we would have to install renewable energy generating facilities EACH YEAR FOR FORTY YEARS equalling 400/11.5 = 35 times as much renewable generating capacity as has ever been installed.

With current technology, the solar electric nameplate capacity of one square meter of solar panel is about 130 watts. Assuming that solar electric panels are installed with wind turbines in the ratio 4:7.5, the area of solar electric panels required each year is 35 x 4 GW/yr / 130 W/m2 = 1e9 m2 = 1000 square kilometres of solar panels per year for forty years. (1000 square kilometres is a square 32 kilometres on a side).

Assuming three megawatts per wind turbine (the biggest current designs), we would need to install 35 * 7.5 GW/yr / 3MW = 87,000 wind turbines per year for forty years.

To keep the sum of wind electricity power, solar electricity power, and oil power flat during the forty year decline of oil production would require 1000 square kilometres of solar panels and 87,000 very big wind turbines to be installed every year of the decline.

David Delaney, Ottawa

I calculated only the new power required each year from wind and solar electricity to keep the total WORLD power from oil, wind, and solar electricity constant. The calculation is conservative (low) for several reasons. First, it makes no allowance for the fact that growth of this total, rather than constancy, may be necessary for the stability of the world economy. Second, it makes no allowance for the need for new infrastructure to support the replacement of oil energy by electricity from renewable resources. The energy requirements of this new infrastructure will surely increase the amount of renewable energy needed. Third, it makes no allowance for the energy losses involved in the creation of liquid fuels for energy storage and transportation. Fourth, it does not contemplate any decrease of production of natural gas, except as represented by the decrease of natural gas liquids, a small part of natural gas energy.

My calculation does not consider the current electricity supply except to the extent that it depends on oil. This is a very small extent indeed, not just in the US, but in the world as a whole. Almost all electricity is provided by coal, natural gas, nuclear fission, and hydro. Also, my calculation is for the world, not for the US.

Why Efficiency Gains Won't Reduce Consumption (undated)

Jeavons' paradox says energy efficiency improvements tend to increase energy consumption, so without other consumption controls, efficiency improvements cannot decrease consumption.

When there are "other consumption controls", for example irremediable depletion or punishing taxes, efficiency improvements mitigate the economic effect of having to make do with less energy, but are limited in the three ways described below. In the case of irremediable depletion, these limitations must be understood in the context of the relationship of energy to economic growth. The economy grows only when more things, or bigger things, or more services, are delivered. At constant energy efficiency, economic growth in any sector requires growing energy use.

Limitation 1) Rate of response. Efficiency gains are made by designing and making new products, buildings, transportation systems, and infrastructure. The design and implementation takes time - years. The expected oil production decline is between two and three percent per year, indefinitely. Actually, it's worse - the decline of maximum possible production is expected to be linear, a constant number of barrels per year at about two percent of the 2010 expected production rate for about forty years. (See the ASPO graph.) This is an enormous rate of decline of the energy input to society. The rate of introduction of efficiency improvements will have great difficulty keeping up with the decline. If growth is to be maintained, the yearly introduction of efficiency improvements must be greater than the rate of decline of energy sources. Less energy will be used year after year, certainly, but at least some of the decrease seems likely to be due to stopping economically productive activity.

Limitation 2) A high rate of introduction of energy efficiency improvements requires a high rate of economic activity and a correspondingly high rate of energy usage. In times of relentlessly decreasing energy availability, there will be severe competition between alternative uses of limited energy - to feed people and heat houses, to provide consumer goods, or to invest in efficiency improvements.

Limitation 3) In any field, there will be limits to the energy efficiencies that can be made. Even if growth has been maintained by improvements in energy efficiency, when improvements to energy efficiency stop, as they must, or otherwise become insufficient to offset the effects of declining energy input, economic growth must stop.

Against "Nine Reasons Not to Panic"

Letter to Editor of the Globe and Mail (May 29 2005)

Re.: Oil price: Nine reasons not to panic, Ebner, Globe and Mail, May 28

Mr Ebner's "Nine reasons not to panic" (sidebar in Saturday's May 28 ROB) misrepresent the coming peak of world oil production. After the peak, the world's yearly production of oil will decrease forever, presenting huge difficulties for the world economy. Surely this fact will make it a turning point in human history and a valid object for serious concern and planning, whether it happens in 2005 or 2036. Mr Ebner uses uncertainty about the date of occurrence of the peak to imply uncertainty about its existence. I challenge Mr Ebner to poll petroleum geologists respected by their profession. I doubt he will find any who deny the peak will occur in that range of dates. I believe most will say it will occur early in the range.

Mr Ebner confuses the results of exploration for new fields with reserve revision in existing fields. Oil discoveries - volumes of oil discovered by wildcat drilling for new fields - have decreased steadily and dramatically since the 1960s. We now find less than one barrel of oil for each four barrels we burn. The decrease is not due solely to reduced exploration. It applies equally to the yield per wildcat drilled, which accounts for the reluctance of the oil companies to explore. See the studies by Wood MacKenzie.

Mr Ebner would like us to believe that large reserves imply large production. That it's not so is the essence of Hubbert's contribution. Mr Ebner's assertion that we have "sufficient reserves" for four decades at "current rates of usage" means the opposite of what Mr Ebner thinks it means. It means that production will be lower than the current rate of production for most of the next four decades, and much lower long before the end of the four decades. Hubbert taught us that it will take forever to produce (extract) the totality of those "sufficient" reserves.

Why Ethanol Can't "Solve" the Fuels Problem (July 02 2005)

I was recently informed by a fuel ethanol enthusiast that the EROEI (Energy Returned On Energy Invested) of ethanol from agricultural products had been greatly increased from 1.38 to 2. He was incredulous, to say the least, when I told him that was not nearly high enough for ethanol to serve as a primary energy source that could keep business-as-usual going after the oil peak. Actually, he accused me of being a supporter of the oil industry, anti-farmer, and a despoiler of the environment, and would not listen (shouted me down in fact) when I tried to explain the realities to him. So, to restore my equilibrium, I am now imposing on you what he refused to listen to.

I show below that ethanol cannot replace the fuel shortages that peak oil will bring, at least not without a very large increase in the total amount of energy we produce. This will be a problem, to say the least, when the energy available from oil and natural gas is declining.

The deficiency of ethanol is its low energy profit ratio. To make the notion of energy profit ratio a little more precise, consider the following definition of EROEI, or Energy Returned On Energy Invested. (I like to pronounce it ee-ro-ee.)

The EROIE of a primary energy technology is the ratio of Energy Returned to Energy invested.

Energy Returned is the amount of energy that an energy producing technology produces for all uses, including further energy production.

Energy Invested is the amount of energy already available for use by society that must be used by the energy producing technology to produce the Energy Returned.

Note that the Energy Invested is not the same as the sum of the energy inputs to the process of operating the energy producing technology. Energy Invested is only that part of the input energy that is already in a form in which it is ready for consumption in society - gasoline, ethanol, diesel fuel, coal ready to burn, et cetera, but not crude oil, sunlight, wind energy, et cetera.

Also note that an EOREI = 1 is the break-even EROEI. Unless a primary energy technology has an EROEI greater than 1, it is obviously useless. A fuels technology might be useful in special circumstances at an EROEI less than one, but it would take more energy from some other source to produce the fuel than the fuel delivered in use.

Let's take the EROEI of the energy derived from the oil industry in the US as a comparison. Robert Kaufman, , calculates it as about ten for extraction in the US in 2000. Because the US is a very mature oil province, this is probably low for the world as a whole. The value ten is read from image 33 in his talk at Lawrence Livermore Labs, "Oil and the American Way of Life: Don't Ask Don't Tell",
(An outstanding talk, by the way.)

So, lets compare two transportation fuel producing technologies, ethanol with an EROEI = 2, and diesel, gasoline, JP4, et cetera with an EREOI = 10. The relevant question is how much energy does society have to produce in total to get the same amount of energy for transportation use from each technology?

Petroleum's EROEI = 10 means that for each ten energy-units of oil-derived fuels you produce you get to keep nine energy units for uses other than fuel production, since you have to put aside one energy unit to produce the next ten units.

Ethanol's EROEI = 2 means that for every two energy units of ethanol you produce, you get to keep only one energy unit for uses other than fuel production since you have to put one energy unit aside to produce the next two energy units. Therefore, to produce nine energy units of ethanol fuel for uses other than fuel production you have to produce nine additional energy units for use in fuel production, for a total of eighteen energy units of total energy production for fuels.

To restate in a general form that does not imply the invested energy is necessarily from the energy technology it's invested in: For nine fuel energy units for uses other than fuel production you have to produce a total of ten energy units if the nine fuel energy units are from oil, and eighteen energy units if the nine fuel energy units are from ethanol. In other words ethanol fuels require a 1.8 times the total energy production for a given fuel-energy for uses other than fuel production compared to petroleum fuels.

Therefore, to replace a given amount, FE, of oil-derived fuel energy by ethanol fuel energy, thus keeping the available fuel energy constant would require INCREASING total energy production devoted to fuels by ((18/9) - (10/9)) x FE = 0.89 x FE.

Consider a policy of keeping the energy available from oil plus ethanol fuels for uses other than fuel production constant by replacing the fuels derived from oil as oil production declines at two percent per year after the peak of oil production. This constant fuel energy replacement policy would require the total energy for fuels produced by society to rise by 0.89 x 2% = 1.78% per year - a doubling time of 39 years.

This additional energy would have to come from renewables, coal, nuclear, perhaps even still more ethanol. It is a gigantic amount. World oil production will decline at a fairly constant rate of 0.5 billion barrels per year for forty years (ASPO). This is approximately 100 gigawatts per year per year. If we assume that two-thirds of this is used as fuel, we would require an increase of total energy production of 0.89 * 2/3 * 100 = 59 gigawatts per year per year just to keep the worlds fuel energy from oil plus ethanol flat. This is the energy production of, for example, 59 big nukes or big coal generating plants. That's equal to the additional energy produced by 59 NEW big nukes or new coal plants each year - just to keep the fuel energy from oil plus ethanol constant.

To INCREASE the transportation fuel energy from oil plus ethanol by E% per year by ethanol production would require an ADDITIONAL 2 x E% increase in the total energy production devoted to fuels. So the total increase per year in total energy production for fuels would be 1.78% per year to offset the 2% decline of oil production plus another 2 x E% per year for the E% increase.

In other words, after the oil peak, a 1% per year increase in oil plus ethanol transportation energy would require 3.78% increase per year in total energy production for fuels - a doubling time of eighteen years - a 2% per year increase in oil plus ethanol transportation fuel energy would require a 5.78% increase per year in total energy production for fuels - a doubling time of twelve years.

Good-bye business-as-usual. Or good-bye biosphere. Or both.

David Delaney

Bill Totten


  • A big piece of the answer, of course, is redefining how we elect tolive. It is not at all clear that energy to transport water from Fiji to be consumed in Europe is a wise choice.

    We ordinary people easily understand the foibles of Fiji water, but do not have the tools to evaluate and select among the other choices.

    Either you with that facility are going to explain the choices and the time frame of the need to undertake them ... or some despot will try.

    Something as simple as incentives to move close to work, paid for by the funds that would have provided more roads, makes a tiny but important contribution.

    No single technology appeas likely to save us. Growth ends eventually anyhow. Time to get moving. Many tiny contributions will be necessary.

    The wisdom to lead this politically difficult and technically challenging transition will come from most unusual people, if it comes at all. You who contribute here may be, or may inform, those who lead. Avoiding panic will be the challenge of this century.

    You have begun the quest; do not weaken!

    You know wha depends on your success...

    By Anonymous Anonymous, at 12:55 PM, August 03, 2007  

Post a Comment

<< Home